------------------------------------------------------------------------------------------------------------------
      name:  <unnamed>
       log:  C:\Users\mbeissin\Desktop\Stata files for book\Robustnesstestfiles\Logfiles\robustnesstestschapter7.l
> og
  log type:  text
 opened on:  26 Jan 2022, 13:53:29

. * =============================================================================
. * Robustness checks for results reported in Chapter 7  
. * Author: Mark R. Beissinger  
. * Date:  January 2022  
. * Princeton, NJ 
. * =============================================================================
. * BEFORE RUNNING, YOU MUST SET THE DEFAULT PATH FOR WHERE THE DATA
. *   FILES RESIDE
. * ============================================================================
. * The following datafiles are used in this chapter:
. *   Monitoring surveys (Ukraine)--monitoring20052014engmerged.dta
. *   KIIS survey (Ukraine)--mohyla.orangerev.dta
. *   2011 Arab Barometer--Tunisia and Egypt--fullarabbarom2.dta
. *   NOTE: The Gallup data cited in chapter are not available for replication; readers 
. *         interested in these data must consult with the Gallup Organization directly
. * =============================================================================
. * Before running, download the following packages for STATA:
. * checkrob from http://fmwww.bc.edu/RePEc/bocode/c
. * spost_13.ado from https://jslsoc.sitehost.iu.edu/spost13.htm for Brant test
. * smhsiao module from http://fmwww.bc.edu/RePEc/bocode/s
. * collin from https://stats.oarc.ucla.edu/stata/ado/analysis/
. * =============================================================================
. * The following output is generated by these robustness tests:
. *       Robustnesstestfiles\Logfiles\robustnesstestschapter7.log
. *       Excel file with the reconfigured results of checkrob tests
. *               Robustnesstestfiles\Outputfiles\checkrob.results.chapter7.xlsx
. * =============================================================================
. 
. 
. * =================================================================
. * TESTING BASE MODELS FOR PROPORTIONAL ODDS ASSUMPTION--BRANT TEST
. * =================================================================
. 
. * Ukraine 2005
. clear

. use monitoring20052014engmerged.dta

. * Testing the proportional odds assumption of ordered logit on gender and age base model
. ologit partic5a gender newage if EVA_vers=="yr2005", or

Iteration 0:   log likelihood = -2367.5224  
Iteration 1:   log likelihood = -2360.9553  
Iteration 2:   log likelihood = -2360.9526  
Iteration 3:   log likelihood = -2360.9526  

Ordered logistic regression                     Number of obs     =      1,747
                                                LR chi2(2)        =      13.14
                                                Prob > chi2       =     0.0014
Log likelihood = -2360.9526                     Pseudo R2         =     0.0028

------------------------------------------------------------------------------
    partic5a | Odds Ratio   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      gender |   .7916976   .0702516    -2.63   0.008     .6653154    .9420871
      newage |   1.005816   .0026195     2.23   0.026     1.000695    1.010963
-------------+----------------------------------------------------------------
       /cut1 |  -1.501182   .1447168                     -1.784822   -1.217542
       /cut2 |  -1.269646   .1430481                     -1.550015   -.9892769
       /cut3 |   .4325869   .1381726                      .1617736    .7034002
       /cut4 |   .8040097    .138588                      .5323821    1.075637
------------------------------------------------------------------------------

. brant, detail

Estimated coefficients from binary logits

----------------------------------------------------------
    Variable |  y_gt_1     y_gt_2     y_gt_3     y_gt_4   
-------------+--------------------------------------------
      gender |   -0.441     -0.430     -0.199     -0.094  
             |    -3.31      -3.49      -2.04      -0.92  
      newage |    0.025      0.020     -0.004      0.004  
             |     6.07       5.26      -1.37       1.28  
       _cons |    0.785      0.766      0.004     -0.770  
             |     3.96       4.13       0.03      -4.90  
----------------------------------------------------------
                                               legend: b/t

Brant test of parallel regression assumption

              |       chi2     p>chi2      df
 -------------+------------------------------
          All |      76.82      0.000       6
 -------------+------------------------------
       gender |       7.69      0.053       3
       newage |      69.10      0.000       3

A significant test statistic provides evidence that the parallel
regression assumption has been violated.

. *       RESULT:  Test failed--must use multinomial logit
. 
. * Ukraine 2014
. ologit partic5a gender newage if EVA_vers=="yr2014", or

Iteration 0:   log likelihood = -2440.8253  
Iteration 1:   log likelihood = -2433.0337  
Iteration 2:   log likelihood =  -2433.029  
Iteration 3:   log likelihood =  -2433.029  

Ordered logistic regression                     Number of obs     =      1,752
                                                LR chi2(2)        =      15.59
                                                Prob > chi2       =     0.0004
Log likelihood =  -2433.029                     Pseudo R2         =     0.0032

------------------------------------------------------------------------------
    partic5a | Odds Ratio   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      gender |   .7502822   .0666681    -3.23   0.001     .6303606     .893018
      newage |   1.005197   .0027138     1.92   0.055     .9998919     1.01053
-------------+----------------------------------------------------------------
       /cut1 |  -2.320442   .1591327                     -2.632336   -2.008547
       /cut2 |  -1.564046   .1486435                     -1.855382    -1.27271
       /cut3 |   .4091755   .1423814                      .1301131    .6882379
       /cut4 |    .979549   .1437636                      .6977776     1.26132
------------------------------------------------------------------------------

. brant, detail

Estimated coefficients from binary logits

----------------------------------------------------------
    Variable |  y_gt_1     y_gt_2     y_gt_3     y_gt_4   
-------------+--------------------------------------------
      gender |   -0.610     -0.380     -0.320     -0.120  
             |    -3.41      -2.88      -3.26      -1.13  
      newage |    0.016      0.012      0.002      0.004  
             |     2.76       2.91       0.70       1.20  
       _cons |    2.051      1.310     -0.248     -0.987  
             |     7.29       6.26      -1.59      -5.82  
----------------------------------------------------------
                                               legend: b/t

Brant test of parallel regression assumption

              |       chi2     p>chi2      df
 -------------+------------------------------
          All |      19.11      0.004       6
 -------------+------------------------------
       gender |      11.77      0.008       3
       newage |       7.08      0.070       3

A significant test statistic provides evidence that the parallel
regression assumption has been violated.

. *       RESULT:  Test failed--must use multinomial logit
. 
. * Egyptian Revolution
. clear

. use fullarabbarom2.dta

. ologit egpartic5 gender newage if country==2, or

Iteration 0:   log likelihood = -1211.6406  
Iteration 1:   log likelihood = -1201.7426  
Iteration 2:   log likelihood = -1201.6838  
Iteration 3:   log likelihood = -1201.6838  

Ordered logistic regression                     Number of obs     =      1,211
                                                LR chi2(2)        =      19.91
                                                Prob > chi2       =     0.0000
Log likelihood = -1201.6838                     Pseudo R2         =     0.0082

------------------------------------------------------------------------------
   egpartic5 | Odds Ratio   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      gender |   .5762348     .07188    -4.42   0.000     .4512528    .7358325
      newage |   1.003495   .0043847     0.80   0.425     .9949383    1.012126
-------------+----------------------------------------------------------------
       /cut1 |  -2.687374   .2097097                     -3.098397   -2.276351
       /cut2 |  -2.388099   .2033403                     -2.786639   -1.989559
       /cut3 |   1.187669   .1893775                      .8164961    1.558842
       /cut4 |   1.653906   .1935849                      1.274486    2.033325
------------------------------------------------------------------------------

. brant, detail

Estimated coefficients from binary logits

----------------------------------------------------------
    Variable |  y_gt_1     y_gt_2     y_gt_3     y_gt_4   
-------------+--------------------------------------------
      gender |   -1.256     -0.973     -0.345     -0.461  
             |    -4.94      -4.59      -2.42      -2.77  
      newage |    0.018      0.011      0.001     -0.007  
             |     2.12       1.48       0.26      -1.09  
       _cons |    2.628      2.392     -1.188     -1.295  
             |     7.21       7.64      -5.53      -5.21  
----------------------------------------------------------
                                               legend: b/t

Brant test of parallel regression assumption

              |       chi2     p>chi2      df
 -------------+------------------------------
          All |      19.61      0.003       6
 -------------+------------------------------
       gender |      12.00      0.007       3
       newage |       8.27      0.041       3

A significant test statistic provides evidence that the parallel
regression assumption has been violated.

. *       RESULT:  Test failed--must use multinomial logit
. 
. * Tunisian Revolution
. ologit tpartic4 gender newage if country==10, or

Iteration 0:   log likelihood = -1030.2143  
Iteration 1:   log likelihood = -1008.5874  
Iteration 2:   log likelihood = -1008.2433  
Iteration 3:   log likelihood =  -1008.243  
Iteration 4:   log likelihood =  -1008.243  

Ordered logistic regression                     Number of obs     =      1,154
                                                LR chi2(2)        =      43.94
                                                Prob > chi2       =     0.0000
Log likelihood =  -1008.243                     Pseudo R2         =     0.0213

------------------------------------------------------------------------------
    tpartic4 | Odds Ratio   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      gender |   .4524869   .0600633    -5.97   0.000     .3488327    .5869417
      newage |   1.013519   .0041824     3.25   0.001     1.005355     1.02175
-------------+----------------------------------------------------------------
       /cut1 |  -1.650399   .1907804                     -2.024321   -1.276476
       /cut2 |   1.981808    .194451                      1.600691    2.362925
       /cut3 |   3.585299    .239415                      3.116054    4.054543
------------------------------------------------------------------------------

. brant, detail

Estimated coefficients from binary logits

-----------------------------------------------
    Variable |  y_gt_1     y_gt_2     y_gt_3   
-------------+---------------------------------
      gender |   -1.662     -0.097      0.634  
             |    -8.16      -0.58       1.86  
      newage |    0.039     -0.007     -0.026  
             |     6.00      -1.26      -2.17  
       _cons |    1.356     -1.427     -2.748  
             |     4.94      -5.99      -5.62  
-----------------------------------------------
                                    legend: b/t

Brant test of parallel regression assumption

              |       chi2     p>chi2      df
 -------------+------------------------------
          All |      82.54      0.000       4
 -------------+------------------------------
       gender |      49.03      0.000       2
       newage |      38.68      0.000       2

A significant test statistic provides evidence that the parallel
regression assumption has been violated.

. *       RESULT:  Test failed--must use multinomial logit
. 
. * ======================================================
. * TESTING BASE MODELS FOR IIA ASSUMPTION--HAUSMAN TESTS
. * ======================================================
. * Ukraine data
. clear

. use monitoring20052014engmerged.dta

. 
. * UKRAINE 2005 DATA
. quietly: mlogit partic5a gender newage if EVA_vers=="yr2005", rrr b(3)

. estimates store basemod

. * Test excluding 2
. quietly: mlogit partic5a gender newage if EVA_vers=="yr2005" & partic5a~=2, rrr b(3)

. estimates store basemodno2

. hausman basemodno2 basemod, alleqs constant

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |   basemodno2    basemod       Difference          S.E.
-------------+----------------------------------------------------------------
Part_in_rev  |
      gender |    .4103789     .4138332       -.0034543        .0053671
      newage |   -.0321985    -.0323227        .0001243               .
       _cons |    .3700763     .3738193        -.003743        .0081143
-------------+----------------------------------------------------------------
Apathetic    |
      gender |   -.2608113     -.257114       -.0036972         .006703
      newage |   -.0325262    -.0325519        .0000257               .
       _cons |    .1010245     .1007368        .0002877        .0119338
-------------+----------------------------------------------------------------
Oppose_rev   |
      gender |   -.0165963    -.0159172       -.0006791        .0028543
      newage |   -.0081238    -.0081425        .0000188               .
       _cons |    .3194337      .320036       -.0006022               .
------------------------------------------------------------------------------
                          b = consistent under Ho and Ha; obtained from mlogit
           B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test:  Ho:  difference in coefficients not systematic

                  chi2(9) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        0.73
                Prob>chi2 =      0.9998
                (V_b-V_B is not positive definite)

. *  RESULT:  PASSED
. * Test excluding 4
. quietly: mlogit partic5a gender newage if EVA_vers=="yr2005" & partic5a~=4, rrr b(3)

. estimates store basemodno4

. hausman basemodno4 basemod, alleqs constant

Note: the rank of the differenced variance matrix (8) does not equal the number of coefficients being tested (9);
        be sure this is what you expect, or there may be problems computing the test.  Examine the output of your
        estimators for anything unexpected and possibly consider scaling your variables so that the coefficients
        are on a similar scale.

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |   basemodno4    basemod       Difference          S.E.
-------------+----------------------------------------------------------------
Part_in_rev  |
      gender |    .4144012     .4138332         .000568        .0079762
      newage |   -.0327802    -.0323227       -.0004575        .0005937
       _cons |    .3937539     .3738193        .0199345        .0248802
-------------+----------------------------------------------------------------
Aid_rev      |
      gender |    .2941268     .2940644        .0000624               .
      newage |    -.003868    -.0037921       -.0000759        .0009727
       _cons |   -2.368411    -2.372021        .0036099        .0448097
-------------+----------------------------------------------------------------
Oppose_rev   |
      gender |    -.015656    -.0159172        .0002611        .0004782
      newage |   -.0082586    -.0081425        -.000116        .0004093
       _cons |    .3254349      .320036        .0053989        .0190538
------------------------------------------------------------------------------
                          b = consistent under Ho and Ha; obtained from mlogit
           B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test:  Ho:  difference in coefficients not systematic

                  chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        2.98
                Prob>chi2 =      0.9355
                (V_b-V_B is not positive definite)

. *  RESULT:  PASSED
. * Test excluding 5
. quietly: mlogit partic5a gender newage if EVA_vers=="yr2005" & partic5a~=5, rrr b(3)

. estimates store basemodno5

. hausman basemodno5 basemod, alleqs constant

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |   basemodno5    basemod       Difference          S.E.
-------------+----------------------------------------------------------------
Part_in_rev  |
      gender |    .3972768     .4138332       -.0165564        .0196599
      newage |   -.0331716    -.0323227       -.0008489        .0010553
       _cons |    .4189885     .3738193        .0451691        .0488244
-------------+----------------------------------------------------------------
Aid_rev      |
      gender |    .2919729     .2940644       -.0020915         .017722
      newage |   -.0037489    -.0037921        .0000432        .0015556
       _cons |   -2.373131    -2.372021       -.0011101        .0882168
-------------+----------------------------------------------------------------
Apathetic    |
      gender |   -.2741566     -.257114       -.0170425        .0212948
      newage |   -.0335906    -.0325519       -.0010387        .0011522
       _cons |    .1536799     .1007368        .0529431        .0550242
------------------------------------------------------------------------------
                          b = consistent under Ho and Ha; obtained from mlogit
           B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test:  Ho:  difference in coefficients not systematic

                  chi2(9) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        1.07
                Prob>chi2 =      0.9993
                (V_b-V_B is not positive definite)

. *  RESULT:  PASSED
. * Test excluding 3, with base 4
. quietly: mlogit partic5a gender newage if EVA_vers=="yr2005", rrr b(4)

. estimates store altbasemod

. quietly: mlogit partic5a gender newage if EVA_vers=="yr2005" & partic5a~=3, rrr b(4)

. estimates store altbasemodno3

. hausman altbasemodno3 altbasemod, alleqs constant

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |  altbasemod~3  altbasemod     Difference          S.E.
-------------+----------------------------------------------------------------
Part_in_rev  |
      gender |    .6703995     .6709472       -.0005477               .
      newage |   -.0000628     .0002292       -.0002919               .
       _cons |    .2849455     .2730826        .0118629               .
-------------+----------------------------------------------------------------
Aid_rev      |
      gender |    .5327959     .5511784       -.0183825               .
      newage |      .02756     .0287598       -.0011998               .
       _cons |   -2.412399    -2.472758        .0603586               .
-------------+----------------------------------------------------------------
Oppose_rev   |
      gender |    .2264821     .2411969       -.0147147        .0115677
      newage |    .0235101     .0244094       -.0008993               .
       _cons |    .2638858     .2192992        .0445866               .
------------------------------------------------------------------------------
                          b = consistent under Ho and Ha; obtained from mlogit
           B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test:  Ho:  difference in coefficients not systematic

                  chi2(9) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        2.82
                Prob>chi2 =      0.9710
                (V_b-V_B is not positive definite)

. *  RESULT:  PASSED
. drop _est_basemod _est_basemodno2 _est_basemodno4 _est_basemodno5 _est_altbasemod _est_altbasemodno3

. 
. * UKRAINE 2014 DATA
. quietly: mlogit partic5a gender newage if EVA_vers=="yr2014", rrr b(3)

. estimates store basemod

. * Test excluding 2
. quietly: mlogit partic5a gender newage if EVA_vers=="yr2014" & partic5a~=2, rrr b(3)

. estimates store basemodno2

. hausman basemodno2 basemod, alleqs constant

Note: the rank of the differenced variance matrix (8) does not equal the number of coefficients being tested (9);
        be sure this is what you expect, or there may be problems computing the test.  Examine the output of your
        estimators for anything unexpected and possibly consider scaling your variables so that the coefficients
        are on a similar scale.

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |   basemodno2    basemod       Difference          S.E.
-------------+----------------------------------------------------------------
Part_in_rev  |
      gender |    .4956757      .494411        .0012647               .
      newage |   -.0167093    -.0168753         .000166               .
       _cons |   -1.148877    -1.140897       -.0079809               .
-------------+----------------------------------------------------------------
Apathetic    |
      gender |   -.4858877    -.4863427         .000455               .
      newage |   -.0047718    -.0048471        .0000753               .
       _cons |   -.7422626    -.7386221       -.0036405               .
-------------+----------------------------------------------------------------
Oppose_rev   |
      gender |   -.1497507    -.1497551        4.48e-06               .
      newage |    .0002004     .0001959        4.54e-06               .
       _cons |   -.2737366    -.2735238       -.0002129               .
------------------------------------------------------------------------------
                          b = consistent under Ho and Ha; obtained from mlogit
           B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test:  Ho:  difference in coefficients not systematic

                  chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        0.09
                Prob>chi2 =      1.0000
                (V_b-V_B is not positive definite)

. *  RESULT:  PASSED
. * Test excluding 4
. quietly: mlogit partic5a gender newage if EVA_vers=="yr2014" & partic5a~=4, rrr b(3)

. estimates store basemodno4

. hausman basemodno4 basemod, alleqs constant

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |   basemodno4    basemod       Difference          S.E.
-------------+----------------------------------------------------------------
Part_in_rev  |
      gender |    .5017617      .494411        .0073507               .
      newage |   -.0172358    -.0168753       -.0003604        .0005021
       _cons |   -1.128871    -1.140897        .0120253               .
-------------+----------------------------------------------------------------
Aid_rev      |
      gender |     .007409     .0031835        .0042256               .
      newage |     -.00867    -.0086026       -.0000674        .0004797
       _cons |   -1.291823    -1.292917        .0010931               .
-------------+----------------------------------------------------------------
Oppose_rev   |
      gender |   -.1496594    -.1497551        .0000958               .
      newage |    .0002737     .0001959        .0000778        .0002784
       _cons |   -.2771793    -.2735238       -.0036556               .
------------------------------------------------------------------------------
                          b = consistent under Ho and Ha; obtained from mlogit
           B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test:  Ho:  difference in coefficients not systematic

                  chi2(9) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        2.46
                Prob>chi2 =      0.9819
                (V_b-V_B is not positive definite)

. *  RESULT:  PASSED
. * Test excluding 5
. quietly: mlogit partic5a gender newage if EVA_vers=="yr2014" & partic5a~=5, rrr b(3)

. estimates store basemodno5

. hausman basemodno5 basemod, alleqs constant

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |   basemodno5    basemod       Difference          S.E.
-------------+----------------------------------------------------------------
Part_in_rev  |
      gender |    .4863295      .494411       -.0080815        .0126112
      newage |   -.0168785    -.0168753       -3.11e-06        .0007869
       _cons |   -1.136487    -1.140897        .0044092         .044532
-------------+----------------------------------------------------------------
Aid_rev      |
      gender |   -.0014554     .0031835       -.0046388        .0116927
      newage |   -.0087311    -.0086026       -.0001284        .0007526
       _cons |   -1.284951    -1.292917        .0079653        .0439669
-------------+----------------------------------------------------------------
Apathetic    |
      gender |   -.4897755    -.4863427       -.0034328        .0096317
      newage |   -.0051469    -.0048471       -.0002999        .0006662
       _cons |   -.7234628    -.7386221        .0151593        .0373846
------------------------------------------------------------------------------
                          b = consistent under Ho and Ha; obtained from mlogit
           B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test:  Ho:  difference in coefficients not systematic

                  chi2(9) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        0.63
                Prob>chi2 =      0.9999
                (V_b-V_B is not positive definite)

. *  RESULT:  PASSED
. * Test excluding 3, with base 4
. quietly: mlogit partic5a gender newage if EVA_vers=="yr2014", rrr b(4)

. estimates store altbasemod

. quietly: mlogit partic5a gender newage if EVA_vers=="yr2014" & partic5a~=3, rrr b(4)

. estimates store altbasemodno3

. hausman altbasemodno3 altbasemod, alleqs constant

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |  altbasemod~3  altbasemod     Difference          S.E.
-------------+----------------------------------------------------------------
Part_in_rev  |
      gender |    .9779485     .9807537       -.0028052        .0105001
      newage |   -.0118722    -.0120283        .0001561         .000948
       _cons |   -.4077617    -.4022745       -.0054872        .0468628
-------------+----------------------------------------------------------------
Aid_rev      |
      gender |    .4887808     .4895261       -.0007453         .009485
      newage |   -.0036595    -.0037556        .0000961        .0007636
       _cons |   -.5582982    -.5542945       -.0040037        .0414749
-------------+----------------------------------------------------------------
Oppose_rev   |
      gender |    .3385868     .3365875        .0019992        .0082577
      newage |    .0051858      .005043        .0001428        .0005534
       _cons |    .4577563     .4650983        -.007342        .0310166
------------------------------------------------------------------------------
                          b = consistent under Ho and Ha; obtained from mlogit
           B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test:  Ho:  difference in coefficients not systematic

                  chi2(9) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        0.12
                Prob>chi2 =      1.0000
                (V_b-V_B is not positive definite)

. *  RESULT:  PASSED
. drop _est_basemod _est_basemodno2 _est_basemodno4 _est_basemodno5 _est_altbasemod _est_altbasemodno3

. 
. * Egyptian and Tunisian data
. clear

. use fullarabbarom2.dta

. 
. * EGYPTIAN REVOLUTION
. quietly: mlogit egpartic5 gender newage if country==2, rrr b(3)

. estimates store basemod

. * Test excluding 2
. quietly: mlogit egpartic5 gender newage if country==2 & egpartic5~=2, rrr b(3)

. estimates store basemodno2

. hausman basemodno2 basemod, alleqs constant

Note: the rank of the differenced variance matrix (8) does not equal the number of coefficients being tested (9);
        be sure this is what you expect, or there may be problems computing the test.  Examine the output of your
        estimators for anything unexpected and possibly consider scaling your variables so that the coefficients
        are on a similar scale.

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |   basemodno2    basemod       Difference          S.E.
-------------+----------------------------------------------------------------
Participate  |
      gender |    1.204323      1.20254        .0017829        .0072332
      newage |   -.0174037    -.0173391       -.0000646               .
       _cons |   -2.317991    -2.319322        .0013313               .
-------------+----------------------------------------------------------------
Apathetic    |
      gender |      .01817     .0186638       -.0004938        .0044984
      newage |    .0132063     .0132921       -.0000859               .
       _cons |   -2.869646    -2.873413         .003767               .
-------------+----------------------------------------------------------------
Oppose       |
      gender |   -.3590202    -.3592947        .0002746        .0044327
      newage |    -.006542    -.0066265        .0000845               .
       _cons |   -1.135205    -1.131803       -.0034027               .
------------------------------------------------------------------------------
                          b = consistent under Ho and Ha; obtained from mlogit
           B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test:  Ho:  difference in coefficients not systematic

                  chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        0.04
                Prob>chi2 =      1.0000
                (V_b-V_B is not positive definite)

. * RESULT:  PASSED
. * Test excluding 4
. quietly: mlogit egpartic5 gender newage if country==2 & egpartic5~=4, rrr b(3)

. estimates store basemodno4

. hausman basemodno4 basemod, alleqs constant

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |   basemodno4    basemod       Difference          S.E.
-------------+----------------------------------------------------------------
Participate  |
      gender |    1.201304      1.20254       -.0012353               .
      newage |   -.0172799    -.0173391        .0000591        .0003919
       _cons |   -2.320801    -2.319322       -.0014787        .0201516
-------------+----------------------------------------------------------------
Aid          |
      gender |    .1275034     .1271603        .0003432               .
      newage |    .0096598      .009627        .0000328        .0006409
       _cons |   -3.852313    -3.850805       -.0015077        .0312706
-------------+----------------------------------------------------------------
Oppose       |
      gender |   -.3595134    -.3592947       -.0002187               .
      newage |   -.0066759    -.0066265       -.0000494        .0003783
       _cons |   -1.129788    -1.131803        .0020146        .0165246
------------------------------------------------------------------------------
                          b = consistent under Ho and Ha; obtained from mlogit
           B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test:  Ho:  difference in coefficients not systematic

                  chi2(9) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        0.14
                Prob>chi2 =      1.0000
                (V_b-V_B is not positive definite)

. * RESULT:  PASSED
. * Test excluding 5
. quietly: mlogit egpartic5 gender newage if country==2 & egpartic5~=5, rrr b(3)

. estimates store basemodno5

. hausman basemodno5 basemod, alleqs constant

Note: the rank of the differenced variance matrix (8) does not equal the number of coefficients being tested (9);
        be sure this is what you expect, or there may be problems computing the test.  Examine the output of your
        estimators for anything unexpected and possibly consider scaling your variables so that the coefficients
        are on a similar scale.

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |   basemodno5    basemod       Difference          S.E.
-------------+----------------------------------------------------------------
Participate  |
      gender |    1.198855      1.20254       -.0036848               .
      newage |   -.0172272    -.0173391        .0001118        .0006833
       _cons |    -2.32116    -2.319322       -.0018385        .0375657
-------------+----------------------------------------------------------------
Aid          |
      gender |    .1273698     .1271603        .0002095               .
      newage |    .0098119      .009627        .0001849        .0017628
       _cons |   -3.858422    -3.850805       -.0076166        .0723846
-------------+----------------------------------------------------------------
Apathetic    |
      gender |    .0187229     .0186638        .0000591        .0056029
      newage |    .0135397     .0132921        .0002476        .0011209
       _cons |   -2.883584    -2.873413       -.0101709        .0455781
------------------------------------------------------------------------------
                          b = consistent under Ho and Ha; obtained from mlogit
           B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test:  Ho:  difference in coefficients not systematic

                  chi2(8) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        0.09
                Prob>chi2 =      1.0000
                (V_b-V_B is not positive definite)

. * RESULT:  PASSED
. * Test excluding 3 (with base 4)
. quietly: mlogit egpartic5 gender newage if country==2, rrr b(4)

. estimates store altbasemod

. quietly: mlogit egpartic5 gender newage if country==2 & egpartic5~=3, rrr b(4)

. estimates store altbasemodno3

. hausman altbasemodno3 altbasemod, alleqs constant

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |  altbasemod~3  altbasemod     Difference          S.E.
-------------+----------------------------------------------------------------
Participate  |
      gender |    1.198811     1.183876        .0149355        .0496817
      newage |   -.0317578    -.0306312       -.0011266        .0028556
       _cons |    .5882909     .5540912        .0341997        .0790469
-------------+----------------------------------------------------------------
Aid          |
      gender |    .1091042     .1084965        .0006077        .0177781
      newage |   -.0037106    -.0036651       -.0000455        .0015056
       _cons |   -.9758012    -.9773922         .001591        .0469946
-------------+----------------------------------------------------------------
Oppose       |
      gender |   -.3729291    -.3779585        .0050294         .026413
      newage |   -.0198089    -.0199186        .0001097        .0012664
       _cons |    1.735112     1.741611       -.0064989        .0501793
------------------------------------------------------------------------------
                          b = consistent under Ho and Ha; obtained from mlogit
           B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test:  Ho:  difference in coefficients not systematic

                  chi2(9) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        0.18
                Prob>chi2 =      1.0000
                (V_b-V_B is not positive definite)

. * RESULT:  PASSED
. drop _est_basemod _est_basemodno2 _est_basemodno4 _est_basemodno5 _est_altbasemod _est_altbasemodno3

. 
. * TUNISIAN REVOLUTION
. quietly: mlogit tpartic4 gender newage if country==10, rrr b(2)

. estimates store basemod

. * Test excluding 3
. * quietly: mlogit tpartic4 gender newage if country==10 & tpartic4~=3, rrr b(2)
. * estimates store basemodno3
. * hausman basemodno3 basemod, alleqs constant
. * Failed to meet the asymptotic assumptions of the Hausman test
. * Must use suest
. * Suest test
. quietly: mlogit tpartic4 gender newage if country==10, rrr b(2)

. estimates store m1

. quietly: mlogit tpartic4 gender newage if country==10 & tpartic4~=3, rrr b(2)

. estimates store m3

. suest m1 m3, noomitted

Simultaneous results for m1, m3

                                                Number of obs     =      1,154

------------------------------------------------------------------------------
             |               Robust
             |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
m1_Partici~e |
      gender |   1.702776   .2044877     8.33   0.000     1.301987    2.103564
      newage |  -.0415961   .0065242    -6.38   0.000    -.0543832   -.0288089
       _cons |  -1.072174   .2837352    -3.78   0.000    -1.628285   -.5160636
-------------+----------------------------------------------------------------
m1_Support   |
-------------+----------------------------------------------------------------
m1_Apathet~e |
      gender |  -.0293111   .1928888    -0.15   0.879    -.4073662    .3487439
      newage |  -.0079609   .0060822    -1.31   0.191    -.0198818    .0039601
       _cons |  -1.481634   .2575494    -5.75   0.000    -1.986422    -.976847
-------------+----------------------------------------------------------------
m1_Oppose    |
      gender |   .9084897   .3508509     2.59   0.010     .2208347    1.596145
      newage |  -.0343097   .0144158    -2.38   0.017     -.062564   -.0060553
       _cons |  -2.226751   .5202866    -4.28   0.000    -3.246494   -1.207008
-------------+----------------------------------------------------------------
m3_Partici~e |
      gender |   1.709258   .2048567     8.34   0.000     1.307747     2.11077
      newage |  -.0412992   .0064087    -6.44   0.000    -.0538601   -.0287384
       _cons |  -1.085901   .2804486    -3.87   0.000     -1.63557   -.5362318
-------------+----------------------------------------------------------------
m3_Support   |
-------------+----------------------------------------------------------------
m3_Oppose    |
      gender |   .9150565    .351261     2.61   0.009     .2265976    1.603516
      newage |  -.0341283   .0142055    -2.40   0.016    -.0619705   -.0062861
       _cons |  -2.236584    .513454    -4.36   0.000    -3.242935   -1.230232
------------------------------------------------------------------------------

. test [m1_Participate = m3_Participate], cons

 ( 1)  [m1_Participate]gender - [m3_Participate]gender = 0
 ( 2)  [m1_Participate]newage - [m3_Participate]newage = 0
 ( 3)  [m1_Participate]_cons - [m3_Participate]_cons = 0

           chi2(  3) =    3.78
         Prob > chi2 =    0.2868

. * RESULT:  PASSED
. * Test excluding 4
. quietly: mlogit tpartic4 gender newage if country==10 & tpartic4~=4, rrr b(2)

. estimates store basemodno4

. hausman basemodno4 basemod, alleqs constant

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |   basemodno4    basemod       Difference          S.E.
-------------+----------------------------------------------------------------
Participate  |
      gender |    1.714288     1.702776        .0115117        .0128891
      newage |   -.0425315    -.0415961       -.0009354        .0009942
       _cons |   -1.044471    -1.072174        .0277035        .0241419
-------------+----------------------------------------------------------------
Apathetic_inactive|
      gender |   -.0274299    -.0293111        .0018813        .0084691
      newage |   -.0080383    -.0079609       -.0000775        .0006677
       _cons |   -1.479325    -1.481634        .0023099        .0197127
------------------------------------------------------------------------------
                          b = consistent under Ho and Ha; obtained from mlogit
           B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test:  Ho:  difference in coefficients not systematic

                  chi2(6) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        1.95
                Prob>chi2 =      0.9244
                (V_b-V_B is not positive definite)

. * RESULT:  PASSED
. * Test excluding 2 (with base 3)
. quietly: mlogit tpartic4 gender newage if country==10, rrr b(3)

. estimates store altbasemod

. quietly: mlogit tpartic4 gender newage if country==10 & tpartic4~=2, rrr b(3)

. estimates store altbasemodno2

. hausman altbasemodno2 altbasemod, alleqs constant

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |  altbasemod~2  altbasemod     Difference          S.E.
-------------+----------------------------------------------------------------
Participate  |
      gender |    1.693949     1.732087        -.038138         .047619
      newage |   -.0312239    -.0336352        .0024113        .0031055
       _cons |    .3412238       .40946       -.0682362        .0968994
-------------+----------------------------------------------------------------
Oppose       |
      gender |    .9036179     .9378009        -.034183        .0310858
      newage |   -.0244338    -.0263488         .001915        .0004293
       _cons |    -.799096    -.7451169       -.0539791        .0606109
------------------------------------------------------------------------------
                          b = consistent under Ho and Ha; obtained from mlogit
           B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test:  Ho:  difference in coefficients not systematic

                  chi2(6) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        0.33
                Prob>chi2 =      0.9994
                (V_b-V_B is not positive definite)

. * RESULT:  PASSED
. drop _est_basemod _est_basemodno4 _est_altbasemod _est_altbasemodno2 _est_m1 _est_m3

. 
. * ==========================================================
. * TESTING BASE MODELS FOR IIA ASSUMPTION--SMALL-HSIAO TESTS
. * ==========================================================
. * set seed to randomly generated number (number previously generated by runiformint(1000000, 9999999))
. * number given here for replicability
. set seed 2667533

. 
. * UKRAINE DATA
. clear

. use monitoring20052014engmerged.dta

. * Generated two randomly selected groups--group2005 and group2014--for use in Small-Hsiao tests
. * gen group2005=rbinomial(1, 0.5) if EVA_vers=="yr2005"
. * gen group2014=rbinomial(1, 0.5) if EVA_vers=="yr2014"
. * Saved groups for replication
. * Performed Small-Hsiao test on 5-part dependent variable for 2005 sample
. smhsiao partic5a age gender if EVA_vers=="yr2005", elim(1) samp(group2005)

Iteration 0:   log likelihood = -2367.5224
Iteration 1:   log likelihood = -2322.4348
Iteration 2:   log likelihood = -2321.6189
Iteration 3:   log likelihood = -2321.6182

Multinomial logistic regression                   Number of obs   =       1747
                                                  LR chi2(8)      =      91.81
                                                  Prob > chi2     =     0.0000
Log likelihood = -2321.6182                       Pseudo R2       =     0.0194

------------------------------------------------------------------------------
    partic5a |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Part in rev  |
         age |  -.0323213   .0045565    -7.09   0.000     -.041252   -.0233906
      gender |   .4134441   .1466226     2.82   0.005     .1260691    .7008191
       _cons |   .3746554   .2230113     1.68   0.093    -.0624387    .8117495
-------------+----------------------------------------------------------------
Aid rev      |
         age |  -.0039624   .0082214    -0.48   0.630     -.020076    .0121512
      gender |   .2933969   .2756349     1.06   0.287    -.2468376    .8336313
       _cons |  -2.363424   .4467353    -5.29   0.000    -3.239009   -1.487839
-------------+----------------------------------------------------------------
Apathetic    |
         age |  -.0324462   .0056377    -5.76   0.000    -.0434959   -.0213966
      gender |  -.2572978   .1860051    -1.38   0.167    -.6218611    .1072655
       _cons |   .0973343   .2663869     0.37   0.715    -.4247744     .619443
-------------+----------------------------------------------------------------
Oppose rev   |
         age |  -.0081656   .0033795    -2.42   0.016    -.0147893   -.0015419
      gender |  -.0162062   .1149385    -0.14   0.888    -.2414815    .2090691
       _cons |   .3216267   .1811749     1.78   0.076    -.0334695     .676723
------------------------------------------------------------------------------
(partic5a==Support rev is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -8.1e+02
  LnL based on restricted model (partic5a==1 omitted):        -8.1e+02
  Test chi2(3) =  5.722, Pr = 0.1260

. smhsiao partic5a age gender if EVA_vers=="yr2005", elim(2) samp(group2005)

Iteration 0:   log likelihood = -2367.5224
Iteration 1:   log likelihood = -2322.4348
Iteration 2:   log likelihood = -2321.6189
Iteration 3:   log likelihood = -2321.6182

Multinomial logistic regression                   Number of obs   =       1747
                                                  LR chi2(8)      =      91.81
                                                  Prob > chi2     =     0.0000
Log likelihood = -2321.6182                       Pseudo R2       =     0.0194

------------------------------------------------------------------------------
    partic5a |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Part in rev  |
         age |  -.0323213   .0045565    -7.09   0.000     -.041252   -.0233906
      gender |   .4134441   .1466226     2.82   0.005     .1260691    .7008191
       _cons |   .3746554   .2230113     1.68   0.093    -.0624387    .8117495
-------------+----------------------------------------------------------------
Aid rev      |
         age |  -.0039624   .0082214    -0.48   0.630     -.020076    .0121512
      gender |   .2933969   .2756349     1.06   0.287    -.2468376    .8336313
       _cons |  -2.363424   .4467353    -5.29   0.000    -3.239009   -1.487839
-------------+----------------------------------------------------------------
Apathetic    |
         age |  -.0324462   .0056377    -5.76   0.000    -.0434959   -.0213966
      gender |  -.2572978   .1860051    -1.38   0.167    -.6218611    .1072655
       _cons |   .0973343   .2663869     0.37   0.715    -.4247744     .619443
-------------+----------------------------------------------------------------
Oppose rev   |
         age |  -.0081656   .0033795    -2.42   0.016    -.0147893   -.0015419
      gender |  -.0162062   .1149385    -0.14   0.888    -.2414815    .2090691
       _cons |   .3216267   .1811749     1.78   0.076    -.0334695     .676723
------------------------------------------------------------------------------
(partic5a==Support rev is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -1.1e+03
  LnL based on restricted model (partic5a==2 omitted):        -1.0e+03
  Test chi2(3) =  2.911, Pr = 0.4056

. smhsiao partic5a age gender if EVA_vers=="yr2005", elim(3) samp(group2005)

Iteration 0:   log likelihood = -2367.5224
Iteration 1:   log likelihood = -2322.4348
Iteration 2:   log likelihood = -2321.6189
Iteration 3:   log likelihood = -2321.6182

Multinomial logistic regression                   Number of obs   =       1747
                                                  LR chi2(8)      =      91.81
                                                  Prob > chi2     =     0.0000
Log likelihood = -2321.6182                       Pseudo R2       =     0.0194

------------------------------------------------------------------------------
    partic5a |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Part in rev  |
         age |  -.0323213   .0045565    -7.09   0.000     -.041252   -.0233906
      gender |   .4134441   .1466226     2.82   0.005     .1260691    .7008191
       _cons |   .3746554   .2230113     1.68   0.093    -.0624387    .8117495
-------------+----------------------------------------------------------------
Aid rev      |
         age |  -.0039624   .0082214    -0.48   0.630     -.020076    .0121512
      gender |   .2933969   .2756349     1.06   0.287    -.2468376    .8336313
       _cons |  -2.363424   .4467353    -5.29   0.000    -3.239009   -1.487839
-------------+----------------------------------------------------------------
Apathetic    |
         age |  -.0324462   .0056377    -5.76   0.000    -.0434959   -.0213966
      gender |  -.2572978   .1860051    -1.38   0.167    -.6218611    .1072655
       _cons |   .0973343   .2663869     0.37   0.715    -.4247744     .619443
-------------+----------------------------------------------------------------
Oppose rev   |
         age |  -.0081656   .0033795    -2.42   0.016    -.0147893   -.0015419
      gender |  -.0162062   .1149385    -0.14   0.888    -.2414815    .2090691
       _cons |   .3216267   .1811749     1.78   0.076    -.0334695     .676723
------------------------------------------------------------------------------
(partic5a==Support rev is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -6.0e+02
  LnL based on restricted model (partic5a==3 omitted):        -6.0e+02
  Test chi2(3) =  5.523, Pr = 0.1373

. smhsiao partic5a age gender if EVA_vers=="yr2005", elim(4) samp(group2005)

Iteration 0:   log likelihood = -2367.5224
Iteration 1:   log likelihood = -2322.4348
Iteration 2:   log likelihood = -2321.6189
Iteration 3:   log likelihood = -2321.6182

Multinomial logistic regression                   Number of obs   =       1747
                                                  LR chi2(8)      =      91.81
                                                  Prob > chi2     =     0.0000
Log likelihood = -2321.6182                       Pseudo R2       =     0.0194

------------------------------------------------------------------------------
    partic5a |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Part in rev  |
         age |  -.0323213   .0045565    -7.09   0.000     -.041252   -.0233906
      gender |   .4134441   .1466226     2.82   0.005     .1260691    .7008191
       _cons |   .3746554   .2230113     1.68   0.093    -.0624387    .8117495
-------------+----------------------------------------------------------------
Aid rev      |
         age |  -.0039624   .0082214    -0.48   0.630     -.020076    .0121512
      gender |   .2933969   .2756349     1.06   0.287    -.2468376    .8336313
       _cons |  -2.363424   .4467353    -5.29   0.000    -3.239009   -1.487839
-------------+----------------------------------------------------------------
Apathetic    |
         age |  -.0324462   .0056377    -5.76   0.000    -.0434959   -.0213966
      gender |  -.2572978   .1860051    -1.38   0.167    -.6218611    .1072655
       _cons |   .0973343   .2663869     0.37   0.715    -.4247744     .619443
-------------+----------------------------------------------------------------
Oppose rev   |
         age |  -.0081656   .0033795    -2.42   0.016    -.0147893   -.0015419
      gender |  -.0162062   .1149385    -0.14   0.888    -.2414815    .2090691
       _cons |   .3216267   .1811749     1.78   0.076    -.0334695     .676723
------------------------------------------------------------------------------
(partic5a==Support rev is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -9.3e+02
  LnL based on restricted model (partic5a==4 omitted):        -9.3e+02
  Test chi2(3) =  4.396, Pr = 0.2218

. smhsiao partic5a age gender if EVA_vers=="yr2005", elim(5) samp(group2005)

Iteration 0:   log likelihood = -2367.5224
Iteration 1:   log likelihood = -2322.4348
Iteration 2:   log likelihood = -2321.6189
Iteration 3:   log likelihood = -2321.6182

Multinomial logistic regression                   Number of obs   =       1747
                                                  LR chi2(8)      =      91.81
                                                  Prob > chi2     =     0.0000
Log likelihood = -2321.6182                       Pseudo R2       =     0.0194

------------------------------------------------------------------------------
    partic5a |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Part in rev  |
         age |  -.0323213   .0045565    -7.09   0.000     -.041252   -.0233906
      gender |   .4134441   .1466226     2.82   0.005     .1260691    .7008191
       _cons |   .3746554   .2230113     1.68   0.093    -.0624387    .8117495
-------------+----------------------------------------------------------------
Aid rev      |
         age |  -.0039624   .0082214    -0.48   0.630     -.020076    .0121512
      gender |   .2933969   .2756349     1.06   0.287    -.2468376    .8336313
       _cons |  -2.363424   .4467353    -5.29   0.000    -3.239009   -1.487839
-------------+----------------------------------------------------------------
Apathetic    |
         age |  -.0324462   .0056377    -5.76   0.000    -.0434959   -.0213966
      gender |  -.2572978   .1860051    -1.38   0.167    -.6218611    .1072655
       _cons |   .0973343   .2663869     0.37   0.715    -.4247744     .619443
-------------+----------------------------------------------------------------
Oppose rev   |
         age |  -.0081656   .0033795    -2.42   0.016    -.0147893   -.0015419
      gender |  -.0162062   .1149385    -0.14   0.888    -.2414815    .2090691
       _cons |   .3216267   .1811749     1.78   0.076    -.0334695     .676723
------------------------------------------------------------------------------
(partic5a==Support rev is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -6.1e+02
  LnL based on restricted model (partic5a==5 omitted):        -6.1e+02
  Test chi2(3) =  5.504, Pr = 0.1384

. * RESULT:  Passed
. * Performed Small-Hsiao test on 5-part dependent variable for 2014 sample
. smhsiao partic5a age gender if EVA_vers=="yr2014", elim(1) samp(group2014)

Iteration 0:   log likelihood = -2440.8253
Iteration 1:   log likelihood = -2423.5816
Iteration 2:   log likelihood = -2423.2622
Iteration 3:   log likelihood =  -2423.262

Multinomial logistic regression                   Number of obs   =       1752
                                                  LR chi2(8)      =      35.13
                                                  Prob > chi2     =     0.0000
Log likelihood =  -2423.262                       Pseudo R2       =     0.0072

------------------------------------------------------------------------------
    partic5a |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Part in rev  |
         age |  -.0169146   .0058713    -2.88   0.004    -.0284221   -.0054071
      gender |   .4941499   .1870729     2.64   0.008     .1274939     .860806
       _cons |  -1.138628   .2947905    -3.86   0.000    -1.716407   -.5608494
-------------+----------------------------------------------------------------
Aid rev      |
         age |  -.0085936   .0058574    -1.47   0.142    -.0200738    .0028866
      gender |   .0031158   .1888884     0.02   0.987    -.3670988    .3733303
       _cons |  -1.292938   .2982153    -4.34   0.000    -1.877429   -.7084468
-------------+----------------------------------------------------------------
Apathetic    |
         age |   -.004914   .0047161    -1.04   0.297    -.0141574    .0043294
      gender |  -.4866075   .1586145    -3.07   0.002    -.7974861   -.1757288
       _cons |  -.7352536   .2412801    -3.05   0.002    -1.208154   -.2623532
-------------+----------------------------------------------------------------
Oppose rev   |
         age |    .000244    .003544     0.07   0.945    -.0067022    .0071902
      gender |  -.1496025   .1163779    -1.29   0.199    -.3776989    .0784939
       _cons |  -.2758406   .1862993    -1.48   0.139    -.6409806    .0892994
------------------------------------------------------------------------------
(partic5a==Support rev is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -9.6e+02
  LnL based on restricted model (partic5a==1 omitted):        -9.6e+02
  Test chi2(3) =  4.211, Pr = 0.2396

. smhsiao partic5a age gender if EVA_vers=="yr2014", elim(2) samp(group2014)

Iteration 0:   log likelihood = -2440.8253
Iteration 1:   log likelihood = -2423.5816
Iteration 2:   log likelihood = -2423.2622
Iteration 3:   log likelihood =  -2423.262

Multinomial logistic regression                   Number of obs   =       1752
                                                  LR chi2(8)      =      35.13
                                                  Prob > chi2     =     0.0000
Log likelihood =  -2423.262                       Pseudo R2       =     0.0072

------------------------------------------------------------------------------
    partic5a |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Part in rev  |
         age |  -.0169146   .0058713    -2.88   0.004    -.0284221   -.0054071
      gender |   .4941499   .1870729     2.64   0.008     .1274939     .860806
       _cons |  -1.138628   .2947905    -3.86   0.000    -1.716407   -.5608494
-------------+----------------------------------------------------------------
Aid rev      |
         age |  -.0085936   .0058574    -1.47   0.142    -.0200738    .0028866
      gender |   .0031158   .1888884     0.02   0.987    -.3670988    .3733303
       _cons |  -1.292938   .2982153    -4.34   0.000    -1.877429   -.7084468
-------------+----------------------------------------------------------------
Apathetic    |
         age |   -.004914   .0047161    -1.04   0.297    -.0141574    .0043294
      gender |  -.4866075   .1586145    -3.07   0.002    -.7974861   -.1757288
       _cons |  -.7352536   .2412801    -3.05   0.002    -1.208154   -.2623532
-------------+----------------------------------------------------------------
Oppose rev   |
         age |    .000244    .003544     0.07   0.945    -.0067022    .0071902
      gender |  -.1496025   .1163779    -1.29   0.199    -.3776989    .0784939
       _cons |  -.2758406   .1862993    -1.48   0.139    -.6409806    .0892994
------------------------------------------------------------------------------
(partic5a==Support rev is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -9.5e+02
  LnL based on restricted model (partic5a==2 omitted):        -9.5e+02
  Test chi2(3) =  5.323, Pr = 0.1496

. smhsiao partic5a age gender if EVA_vers=="yr2014", elim(3) samp(group2014)

Iteration 0:   log likelihood = -2440.8253
Iteration 1:   log likelihood = -2423.5816
Iteration 2:   log likelihood = -2423.2622
Iteration 3:   log likelihood =  -2423.262

Multinomial logistic regression                   Number of obs   =       1752
                                                  LR chi2(8)      =      35.13
                                                  Prob > chi2     =     0.0000
Log likelihood =  -2423.262                       Pseudo R2       =     0.0072

------------------------------------------------------------------------------
    partic5a |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Part in rev  |
         age |  -.0169146   .0058713    -2.88   0.004    -.0284221   -.0054071
      gender |   .4941499   .1870729     2.64   0.008     .1274939     .860806
       _cons |  -1.138628   .2947905    -3.86   0.000    -1.716407   -.5608494
-------------+----------------------------------------------------------------
Aid rev      |
         age |  -.0085936   .0058574    -1.47   0.142    -.0200738    .0028866
      gender |   .0031158   .1888884     0.02   0.987    -.3670988    .3733303
       _cons |  -1.292938   .2982153    -4.34   0.000    -1.877429   -.7084468
-------------+----------------------------------------------------------------
Apathetic    |
         age |   -.004914   .0047161    -1.04   0.297    -.0141574    .0043294
      gender |  -.4866075   .1586145    -3.07   0.002    -.7974861   -.1757288
       _cons |  -.7352536   .2412801    -3.05   0.002    -1.208154   -.2623532
-------------+----------------------------------------------------------------
Oppose rev   |
         age |    .000244    .003544     0.07   0.945    -.0067022    .0071902
      gender |  -.1496025   .1163779    -1.29   0.199    -.3776989    .0784939
       _cons |  -.2758406   .1862993    -1.48   0.139    -.6409806    .0892994
------------------------------------------------------------------------------
(partic5a==Support rev is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -5.9e+02
  LnL based on restricted model (partic5a==3 omitted):        -5.9e+02
  Test chi2(3) =  3.781, Pr = 0.2861

. smhsiao partic5a age gender if EVA_vers=="yr2014", elim(4) samp(group2014)

Iteration 0:   log likelihood = -2440.8253
Iteration 1:   log likelihood = -2423.5816
Iteration 2:   log likelihood = -2423.2622
Iteration 3:   log likelihood =  -2423.262

Multinomial logistic regression                   Number of obs   =       1752
                                                  LR chi2(8)      =      35.13
                                                  Prob > chi2     =     0.0000
Log likelihood =  -2423.262                       Pseudo R2       =     0.0072

------------------------------------------------------------------------------
    partic5a |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Part in rev  |
         age |  -.0169146   .0058713    -2.88   0.004    -.0284221   -.0054071
      gender |   .4941499   .1870729     2.64   0.008     .1274939     .860806
       _cons |  -1.138628   .2947905    -3.86   0.000    -1.716407   -.5608494
-------------+----------------------------------------------------------------
Aid rev      |
         age |  -.0085936   .0058574    -1.47   0.142    -.0200738    .0028866
      gender |   .0031158   .1888884     0.02   0.987    -.3670988    .3733303
       _cons |  -1.292938   .2982153    -4.34   0.000    -1.877429   -.7084468
-------------+----------------------------------------------------------------
Apathetic    |
         age |   -.004914   .0047161    -1.04   0.297    -.0141574    .0043294
      gender |  -.4866075   .1586145    -3.07   0.002    -.7974861   -.1757288
       _cons |  -.7352536   .2412801    -3.05   0.002    -1.208154   -.2623532
-------------+----------------------------------------------------------------
Oppose rev   |
         age |    .000244    .003544     0.07   0.945    -.0067022    .0071902
      gender |  -.1496025   .1163779    -1.29   0.199    -.3776989    .0784939
       _cons |  -.2758406   .1862993    -1.48   0.139    -.6409806    .0892994
------------------------------------------------------------------------------
(partic5a==Support rev is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -8.5e+02
  LnL based on restricted model (partic5a==4 omitted):        -8.5e+02
  Test chi2(3) =  4.398, Pr = 0.2216

. smhsiao partic5a age gender if EVA_vers=="yr2014", elim(5) samp(group2014)

Iteration 0:   log likelihood = -2440.8253
Iteration 1:   log likelihood = -2423.5816
Iteration 2:   log likelihood = -2423.2622
Iteration 3:   log likelihood =  -2423.262

Multinomial logistic regression                   Number of obs   =       1752
                                                  LR chi2(8)      =      35.13
                                                  Prob > chi2     =     0.0000
Log likelihood =  -2423.262                       Pseudo R2       =     0.0072

------------------------------------------------------------------------------
    partic5a |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Part in rev  |
         age |  -.0169146   .0058713    -2.88   0.004    -.0284221   -.0054071
      gender |   .4941499   .1870729     2.64   0.008     .1274939     .860806
       _cons |  -1.138628   .2947905    -3.86   0.000    -1.716407   -.5608494
-------------+----------------------------------------------------------------
Aid rev      |
         age |  -.0085936   .0058574    -1.47   0.142    -.0200738    .0028866
      gender |   .0031158   .1888884     0.02   0.987    -.3670988    .3733303
       _cons |  -1.292938   .2982153    -4.34   0.000    -1.877429   -.7084468
-------------+----------------------------------------------------------------
Apathetic    |
         age |   -.004914   .0047161    -1.04   0.297    -.0141574    .0043294
      gender |  -.4866075   .1586145    -3.07   0.002    -.7974861   -.1757288
       _cons |  -.7352536   .2412801    -3.05   0.002    -1.208154   -.2623532
-------------+----------------------------------------------------------------
Oppose rev   |
         age |    .000244    .003544     0.07   0.945    -.0067022    .0071902
      gender |  -.1496025   .1163779    -1.29   0.199    -.3776989    .0784939
       _cons |  -.2758406   .1862993    -1.48   0.139    -.6409806    .0892994
------------------------------------------------------------------------------
(partic5a==Support rev is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -6.6e+02
  LnL based on restricted model (partic5a==5 omitted):        -6.6e+02
  Test chi2(3) =  3.108, Pr = 0.3752

. * RESULT:  Passed
.  
. * ARAB BAROMETER DATA
. clear

. use fullarabbarom2.dta

. * Small-Hsiao tests
. * set seed to randomly generated number (number generated by runiformint(1000000, 9999999))
. set seed 2667533

. * Generated two randomly selected groups--groupeg and grouptu--for use in Small-Hsiao tests
. * gen groupeg=rbinomial(1, 0.5) if country==2
. * gen grouptu=rbinomial(1, 0.5) if country==10
. * Saved groups for replication
. * Performed Small-Hsiao test on 5-part dependent variable for Egyptian sample
. smhsiao egpartic5 gender newage if country==2, elim(1) samp(groupeg)

Iteration 0:   log likelihood = -1211.6406
Iteration 1:   log likelihood = -1191.9286
Iteration 2:   log likelihood = -1190.9384
Iteration 3:   log likelihood = -1190.9333
Iteration 4:   log likelihood = -1190.9333

Multinomial logistic regression                   Number of obs   =       1211
                                                  LR chi2(8)      =      41.41
                                                  Prob > chi2     =     0.0000
Log likelihood = -1190.9333                       Pseudo R2       =     0.0171

------------------------------------------------------------------------------
   egpartic5 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      gender |    1.20254   .2565398     4.69   0.000      .699731    1.705348
      newage |  -.0173391   .0084208    -2.06   0.039    -.0338435   -.0008346
       _cons |  -2.319322   .3688739    -6.29   0.000    -3.042301   -1.596342
-------------+----------------------------------------------------------------
Aid          |
      gender |   .1271603   .3886634     0.33   0.744    -.6346061    .8889266
      newage |    .009627   .0134621     0.72   0.475    -.0167583    .0360123
       _cons |  -3.850805   .6026895    -6.39   0.000    -5.032055   -2.669556
-------------+----------------------------------------------------------------
Apathetic    |
      gender |   .0186638   .2339147     0.08   0.936    -.4398005    .4771281
      newage |   .0132921   .0080827     1.64   0.100    -.0025497    .0291339
       _cons |  -2.873413   .3639679    -7.89   0.000    -3.586777   -2.160049
-------------+----------------------------------------------------------------
Oppose       |
      gender |  -.3592947   .1695254    -2.12   0.034    -.6915584   -.0270311
      newage |  -.0066265   .0061659    -1.07   0.283    -.0187115    .0054585
       _cons |  -1.131803   .2526693    -4.48   0.000    -1.627025   -.6365797
------------------------------------------------------------------------------
(egpartic5==Support is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -4.4e+02
  LnL based on restricted model (egpartic5==1 omitted):       -4.4e+02
  Test chi2(3) =  4.967, Pr = 0.1742

. smhsiao egpartic5 gender newage if country==2, elim(2) samp(groupeg)

Iteration 0:   log likelihood = -1211.6406
Iteration 1:   log likelihood = -1191.9286
Iteration 2:   log likelihood = -1190.9384
Iteration 3:   log likelihood = -1190.9333
Iteration 4:   log likelihood = -1190.9333

Multinomial logistic regression                   Number of obs   =       1211
                                                  LR chi2(8)      =      41.41
                                                  Prob > chi2     =     0.0000
Log likelihood = -1190.9333                       Pseudo R2       =     0.0171

------------------------------------------------------------------------------
   egpartic5 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      gender |    1.20254   .2565398     4.69   0.000      .699731    1.705348
      newage |  -.0173391   .0084208    -2.06   0.039    -.0338435   -.0008346
       _cons |  -2.319322   .3688739    -6.29   0.000    -3.042301   -1.596342
-------------+----------------------------------------------------------------
Aid          |
      gender |   .1271603   .3886634     0.33   0.744    -.6346061    .8889266
      newage |    .009627   .0134621     0.72   0.475    -.0167583    .0360123
       _cons |  -3.850805   .6026895    -6.39   0.000    -5.032055   -2.669556
-------------+----------------------------------------------------------------
Apathetic    |
      gender |   .0186638   .2339147     0.08   0.936    -.4398005    .4771281
      newage |   .0132921   .0080827     1.64   0.100    -.0025497    .0291339
       _cons |  -2.873413   .3639679    -7.89   0.000    -3.586777   -2.160049
-------------+----------------------------------------------------------------
Oppose       |
      gender |  -.3592947   .1695254    -2.12   0.034    -.6915584   -.0270311
      newage |  -.0066265   .0061659    -1.07   0.283    -.0187115    .0054585
       _cons |  -1.131803   .2526693    -4.48   0.000    -1.627025   -.6365797
------------------------------------------------------------------------------
(egpartic5==Support is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -5.3e+02
  LnL based on restricted model (egpartic5==2 omitted):       -5.3e+02
  Test chi2(3) =  4.607, Pr = 0.2029

. smhsiao egpartic5 gender newage if country==2, elim(3) samp(groupeg)

Iteration 0:   log likelihood = -1211.6406
Iteration 1:   log likelihood = -1191.9286
Iteration 2:   log likelihood = -1190.9384
Iteration 3:   log likelihood = -1190.9333
Iteration 4:   log likelihood = -1190.9333

Multinomial logistic regression                   Number of obs   =       1211
                                                  LR chi2(8)      =      41.41
                                                  Prob > chi2     =     0.0000
Log likelihood = -1190.9333                       Pseudo R2       =     0.0171

------------------------------------------------------------------------------
   egpartic5 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      gender |    1.20254   .2565398     4.69   0.000      .699731    1.705348
      newage |  -.0173391   .0084208    -2.06   0.039    -.0338435   -.0008346
       _cons |  -2.319322   .3688739    -6.29   0.000    -3.042301   -1.596342
-------------+----------------------------------------------------------------
Aid          |
      gender |   .1271603   .3886634     0.33   0.744    -.6346061    .8889266
      newage |    .009627   .0134621     0.72   0.475    -.0167583    .0360123
       _cons |  -3.850805   .6026895    -6.39   0.000    -5.032055   -2.669556
-------------+----------------------------------------------------------------
Apathetic    |
      gender |   .0186638   .2339147     0.08   0.936    -.4398005    .4771281
      newage |   .0132921   .0080827     1.64   0.100    -.0025497    .0291339
       _cons |  -2.873413   .3639679    -7.89   0.000    -3.586777   -2.160049
-------------+----------------------------------------------------------------
Oppose       |
      gender |  -.3592947   .1695254    -2.12   0.034    -.6915584   -.0270311
      newage |  -.0066265   .0061659    -1.07   0.283    -.0187115    .0054585
       _cons |  -1.131803   .2526693    -4.48   0.000    -1.627025   -.6365797
------------------------------------------------------------------------------
(egpartic5==Support is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -2.2e+02
  LnL based on restricted model (egpartic5==3 omitted):       -2.2e+02
  Test chi2(3) =  3.400, Pr = 0.3340

. smhsiao egpartic5 gender newage if country==2, elim(4) samp(groupeg)

Iteration 0:   log likelihood = -1211.6406
Iteration 1:   log likelihood = -1191.9286
Iteration 2:   log likelihood = -1190.9384
Iteration 3:   log likelihood = -1190.9333
Iteration 4:   log likelihood = -1190.9333

Multinomial logistic regression                   Number of obs   =       1211
                                                  LR chi2(8)      =      41.41
                                                  Prob > chi2     =     0.0000
Log likelihood = -1190.9333                       Pseudo R2       =     0.0171

------------------------------------------------------------------------------
   egpartic5 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      gender |    1.20254   .2565398     4.69   0.000      .699731    1.705348
      newage |  -.0173391   .0084208    -2.06   0.039    -.0338435   -.0008346
       _cons |  -2.319322   .3688739    -6.29   0.000    -3.042301   -1.596342
-------------+----------------------------------------------------------------
Aid          |
      gender |   .1271603   .3886634     0.33   0.744    -.6346061    .8889266
      newage |    .009627   .0134621     0.72   0.475    -.0167583    .0360123
       _cons |  -3.850805   .6026895    -6.39   0.000    -5.032055   -2.669556
-------------+----------------------------------------------------------------
Apathetic    |
      gender |   .0186638   .2339147     0.08   0.936    -.4398005    .4771281
      newage |   .0132921   .0080827     1.64   0.100    -.0025497    .0291339
       _cons |  -2.873413   .3639679    -7.89   0.000    -3.586777   -2.160049
-------------+----------------------------------------------------------------
Oppose       |
      gender |  -.3592947   .1695254    -2.12   0.034    -.6915584   -.0270311
      newage |  -.0066265   .0061659    -1.07   0.283    -.0187115    .0054585
       _cons |  -1.131803   .2526693    -4.48   0.000    -1.627025   -.6365797
------------------------------------------------------------------------------
(egpartic5==Support is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -4.5e+02
  LnL based on restricted model (egpartic5==4 omitted):       -4.4e+02
  Test chi2(3) =  5.777, Pr = 0.1230

. smhsiao egpartic5 gender newage if country==2, elim(5) samp(groupeg)

Iteration 0:   log likelihood = -1211.6406
Iteration 1:   log likelihood = -1191.9286
Iteration 2:   log likelihood = -1190.9384
Iteration 3:   log likelihood = -1190.9333
Iteration 4:   log likelihood = -1190.9333

Multinomial logistic regression                   Number of obs   =       1211
                                                  LR chi2(8)      =      41.41
                                                  Prob > chi2     =     0.0000
Log likelihood = -1190.9333                       Pseudo R2       =     0.0171

------------------------------------------------------------------------------
   egpartic5 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      gender |    1.20254   .2565398     4.69   0.000      .699731    1.705348
      newage |  -.0173391   .0084208    -2.06   0.039    -.0338435   -.0008346
       _cons |  -2.319322   .3688739    -6.29   0.000    -3.042301   -1.596342
-------------+----------------------------------------------------------------
Aid          |
      gender |   .1271603   .3886634     0.33   0.744    -.6346061    .8889266
      newage |    .009627   .0134621     0.72   0.475    -.0167583    .0360123
       _cons |  -3.850805   .6026895    -6.39   0.000    -5.032055   -2.669556
-------------+----------------------------------------------------------------
Apathetic    |
      gender |   .0186638   .2339147     0.08   0.936    -.4398005    .4771281
      newage |   .0132921   .0080827     1.64   0.100    -.0025497    .0291339
       _cons |  -2.873413   .3639679    -7.89   0.000    -3.586777   -2.160049
-------------+----------------------------------------------------------------
Oppose       |
      gender |  -.3592947   .1695254    -2.12   0.034    -.6915584   -.0270311
      newage |  -.0066265   .0061659    -1.07   0.283    -.0187115    .0054585
       _cons |  -1.131803   .2526693    -4.48   0.000    -1.627025   -.6365797
------------------------------------------------------------------------------
(egpartic5==Support is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -3.6e+02
  LnL based on restricted model (egpartic5==5 omitted):       -3.6e+02
  Test chi2(3) =  2.475, Pr = 0.4798

. * RESULT:  Passed
. * Performed Small-Hsiao test on 4-part dependent variable for Tunisian sample
. smhsiao tpartic4 gender newage if country==10, elim(1) samp(grouptu)

Iteration 0:   log likelihood = -1030.2143
Iteration 1:   log likelihood = -970.12806
Iteration 2:   log likelihood = -965.22886
Iteration 3:   log likelihood =  -965.1507
Iteration 4:   log likelihood = -965.15065

Multinomial logistic regression                   Number of obs   =       1154
                                                  LR chi2(6)      =     130.13
                                                  Prob > chi2     =     0.0000
Log likelihood = -965.15065                       Pseudo R2       =     0.0632

------------------------------------------------------------------------------
    tpartic4 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      gender |   1.702776   .2064353     8.25   0.000      1.29817    2.107382
      newage |  -.0415961   .0065759    -6.33   0.000    -.0544845   -.0287076
       _cons |  -1.072174   .2783453    -3.85   0.000    -1.617721   -.5266276
-------------+----------------------------------------------------------------
Apathetic/~e |
      gender |  -.0293111   .1910885    -0.15   0.878    -.4038377    .3452154
      newage |  -.0079609   .0062287    -1.28   0.201    -.0201688    .0042471
       _cons |  -1.481634    .268002    -5.53   0.000    -2.006909   -.9563601
-------------+----------------------------------------------------------------
Oppose       |
      gender |   .9084897   .3440836     2.64   0.008     .2340984    1.582881
      newage |  -.0343097   .0122034    -2.81   0.005    -.0582279   -.0103914
       _cons |  -2.226751   .4900106    -4.54   0.000    -3.187154   -1.266348
------------------------------------------------------------------------------
(tpartic4==Support is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -2.7e+02
  LnL based on restricted model (tpartic4==1 omitted):        -2.7e+02
  Test chi2(3) =  6.223, Pr = 0.1012

. smhsiao tpartic4 gender newage if country==10, elim(2) samp(grouptu)

Iteration 0:   log likelihood = -1030.2143
Iteration 1:   log likelihood = -970.12806
Iteration 2:   log likelihood = -965.22886
Iteration 3:   log likelihood =  -965.1507
Iteration 4:   log likelihood = -965.15065

Multinomial logistic regression                   Number of obs   =       1154
                                                  LR chi2(6)      =     130.13
                                                  Prob > chi2     =     0.0000
Log likelihood = -965.15065                       Pseudo R2       =     0.0632

------------------------------------------------------------------------------
    tpartic4 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      gender |   1.702776   .2064353     8.25   0.000      1.29817    2.107382
      newage |  -.0415961   .0065759    -6.33   0.000    -.0544845   -.0287076
       _cons |  -1.072174   .2783453    -3.85   0.000    -1.617721   -.5266276
-------------+----------------------------------------------------------------
Apathetic/~e |
      gender |  -.0293111   .1910885    -0.15   0.878    -.4038377    .3452154
      newage |  -.0079609   .0062287    -1.28   0.201    -.0201688    .0042471
       _cons |  -1.481634    .268002    -5.53   0.000    -2.006909   -.9563601
-------------+----------------------------------------------------------------
Oppose       |
      gender |   .9084897   .3440836     2.64   0.008     .2340984    1.582881
      newage |  -.0343097   .0122034    -2.81   0.005    -.0582279   -.0103914
       _cons |  -2.226751   .4900106    -4.54   0.000    -3.187154   -1.266348
------------------------------------------------------------------------------
(tpartic4==Support is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -1.6e+02
  LnL based on restricted model (tpartic4==2 omitted):        -1.6e+02
  Test chi2(3) =  6.160, Pr = 0.1041

. smhsiao tpartic4 gender newage if country==10, elim(3) samp(grouptu)

Iteration 0:   log likelihood = -1030.2143
Iteration 1:   log likelihood = -970.12806
Iteration 2:   log likelihood = -965.22886
Iteration 3:   log likelihood =  -965.1507
Iteration 4:   log likelihood = -965.15065

Multinomial logistic regression                   Number of obs   =       1154
                                                  LR chi2(6)      =     130.13
                                                  Prob > chi2     =     0.0000
Log likelihood = -965.15065                       Pseudo R2       =     0.0632

------------------------------------------------------------------------------
    tpartic4 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      gender |   1.702776   .2064353     8.25   0.000      1.29817    2.107382
      newage |  -.0415961   .0065759    -6.33   0.000    -.0544845   -.0287076
       _cons |  -1.072174   .2783453    -3.85   0.000    -1.617721   -.5266276
-------------+----------------------------------------------------------------
Apathetic/~e |
      gender |  -.0293111   .1910885    -0.15   0.878    -.4038377    .3452154
      newage |  -.0079609   .0062287    -1.28   0.201    -.0201688    .0042471
       _cons |  -1.481634    .268002    -5.53   0.000    -2.006909   -.9563601
-------------+----------------------------------------------------------------
Oppose       |
      gender |   .9084897   .3440836     2.64   0.008     .2340984    1.582881
      newage |  -.0343097   .0122034    -2.81   0.005    -.0582279   -.0103914
       _cons |  -2.226751   .4900106    -4.54   0.000    -3.187154   -1.266348
------------------------------------------------------------------------------
(tpartic4==Support is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -2.9e+02
  LnL based on restricted model (tpartic4==3 omitted):        -2.9e+02
  Test chi2(3) =  3.234, Pr = 0.3569

. smhsiao tpartic4 gender newage if country==10, elim(4) samp(grouptu)

Iteration 0:   log likelihood = -1030.2143
Iteration 1:   log likelihood = -970.12806
Iteration 2:   log likelihood = -965.22886
Iteration 3:   log likelihood =  -965.1507
Iteration 4:   log likelihood = -965.15065

Multinomial logistic regression                   Number of obs   =       1154
                                                  LR chi2(6)      =     130.13
                                                  Prob > chi2     =     0.0000
Log likelihood = -965.15065                       Pseudo R2       =     0.0632

------------------------------------------------------------------------------
    tpartic4 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      gender |   1.702776   .2064353     8.25   0.000      1.29817    2.107382
      newage |  -.0415961   .0065759    -6.33   0.000    -.0544845   -.0287076
       _cons |  -1.072174   .2783453    -3.85   0.000    -1.617721   -.5266276
-------------+----------------------------------------------------------------
Apathetic/~e |
      gender |  -.0293111   .1910885    -0.15   0.878    -.4038377    .3452154
      newage |  -.0079609   .0062287    -1.28   0.201    -.0201688    .0042471
       _cons |  -1.481634    .268002    -5.53   0.000    -2.006909   -.9563601
-------------+----------------------------------------------------------------
Oppose       |
      gender |   .9084897   .3440836     2.64   0.008     .2340984    1.582881
      newage |  -.0343097   .0122034    -2.81   0.005    -.0582279   -.0103914
       _cons |  -2.226751   .4900106    -4.54   0.000    -3.187154   -1.266348
------------------------------------------------------------------------------
(tpartic4==Support is the base outcome)

Small-Hsiao Test of IIA
-----------------------

  LnL for weighted average of coeffients in two half-samples: -4.1e+02
  LnL based on restricted model (tpartic4==4 omitted):        -4.1e+02
  Test chi2(3) =  5.797, Pr = 0.1219

. * RESULT:  Passed
. 
. * ===================================================================
. * ROBUSTNESS TESTS FOR FIGURE 7.9: PARTICIPATION BY LANGUAGE USE AT 
. *       HOME (UKRAINE ONLY)
. * ===================================================================
. * Ukraine data
. clear

. use monitoring20052014engmerged.dta

. * Multinomial logit, but examining only the participation portion of the regression
. * Orange Revolution
. * checkrob command
. * checkrob 2 3 ch7fig7_9a.txt: mlogit partic5a gender newage edulevel consumergoods ukrspeakathome if version==0
> , rrr
. *       Reworked results can be found in the Excel file checkrob.results.chapter7.xlsx,
. *               under the Robustnesstests/Outputfiles directory
. *       RESULT:  All patterns of significance stable (with exception of edulevel, which is insignificant when co
> nsumergoods included but ukrspeakathome not included); all signs stable
. * Robust standard errors
. mlogit partic5a  gender newage edulevel consumergoods ukrspeakathome  if EVA_vers=="yr2005", rrr vce(robust)

Iteration 0:   log pseudolikelihood = -2366.4614  
Iteration 1:   log pseudolikelihood = -2037.8904  
Iteration 2:   log pseudolikelihood = -2022.2983  
Iteration 3:   log pseudolikelihood =  -2022.198  
Iteration 4:   log pseudolikelihood =  -2022.198  

Multinomial logistic regression                 Number of obs     =      1,746
                                                Wald chi2(20)     =     522.35
                                                Prob > chi2       =     0.0000
Log pseudolikelihood =  -2022.198               Pseudo R2         =     0.1455

--------------------------------------------------------------------------------
               |               Robust
      partic5a |        RRR   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
Part_in_rev    |
        gender |   1.494754   .2287692     2.63   0.009     1.107376    2.017644
        newage |   .9723597    .004966    -5.49   0.000     .9626752    .9821418
      edulevel |   1.166069   .0794453     2.26   0.024     1.020307    1.332654
 consumergoods |   1.216015   .0419615     5.67   0.000     1.136491    1.301103
ukrspeakathome |   3.137825   .5465924     6.56   0.000     2.230252    4.414724
         _cons |   .1432741   .0596102    -4.67   0.000     .0633898    .3238295
---------------+----------------------------------------------------------------
Aid_rev        |
        gender |   1.335445   .3642297     1.06   0.289     .7824724    2.279203
        newage |   .9974216   .0089747    -0.29   0.774     .9799858    1.015168
      edulevel |    1.02673    .131054     0.21   0.836     .7994787    1.318577
 consumergoods |   1.130487   .0714411     1.94   0.052     .9987896     1.27955
ukrspeakathome |    2.38496    .765427     2.71   0.007     1.271444    4.473681
         _cons |   .0257419   .0206891    -4.55   0.000     .0053274    .1243833
---------------+----------------------------------------------------------------
Support_rev    |  (base outcome)
---------------+----------------------------------------------------------------
Apathetic      |
        gender |   .7300962   .1394366    -1.65   0.100     .5021278    1.061563
        newage |   .9727284   .0060914    -4.42   0.000     .9608624     .984741
      edulevel |   1.001542   .0825179     0.02   0.985     .8521926    1.177065
 consumergoods |   .9202741   .0411492    -1.86   0.063     .8430562    1.004565
ukrspeakathome |   .2301635   .0494259    -6.84   0.000      .151094    .3506111
         _cons |   2.407614   .9924625     2.13   0.033     1.073263     5.40092
---------------+----------------------------------------------------------------
Oppose_rev     |
        gender |   .8689785   .1121545    -1.09   0.277     .6747594    1.119101
        newage |   .9985631   .0038718    -0.37   0.711     .9910032    1.006181
      edulevel |   .8959331   .0539518    -1.82   0.068     .7961914     1.00817
 consumergoods |   1.012041   .0292732     0.41   0.679     .9562627    1.071073
ukrspeakathome |   .0887285   .0136517   -15.74   0.000     .0656293    .1199577
         _cons |   2.714424   .8166488     3.32   0.001     1.505183    4.895152
--------------------------------------------------------------------------------

. *       RESULT:  No changes
. * Bootstrapped standard errors
. mlogit partic5a  gender newage edulevel consumergoods ukrspeakathome if EVA_vers=="yr2005", rrr vce(bootstrap, b
> ca seed(1234) rep(1000))
(running mlogit on estimation sample)

Jackknife replications (1746)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400
..................................................   450
..................................................   500
..................................................   550
..................................................   600
..................................................   650
..................................................   700
..................................................   750
..................................................   800
..................................................   850
..................................................   900
..................................................   950
..................................................  1000
..................................................  1050
..................................................  1100
..................................................  1150
..................................................  1200
..................................................  1250
..................................................  1300
..................................................  1350
..................................................  1400
..................................................  1450
..................................................  1500
..................................................  1550
..................................................  1600
..................................................  1650
..................................................  1700
..............................................

Bootstrap replications (1000)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
.......x..........x...............x..........x....    50
.....x........x...................................   100
x..........x......x................x..........x...   150
......x...........................................   200
........x...........x.....x.......................   250
.x........x.x...........x.........................   300
.....x...x....x.......x.......x...x.........x.....   350
............................x...............x...x.   400
.....................x.....x......................   450
x..............x..................................   500
..x............................x.x...............x   550
.........x.....................x........x...x.x...   600
.................x.........x..x......x.x...x..x...   650
..................................................   700
.................x...x.x..........................   750
....................x.xxx.................x.......   800
.....x............xx.............x....x...........   850
x.............................x...........x.......   900
..............x.xx....x............x....x.........   950
..x.......x...........x....xxx.x............x....x  1000

Multinomial logistic regression                 Number of obs     =      1,746
                                                Replications      =        920
                                                Wald chi2(20)     =     506.08
                                                Prob > chi2       =     0.0000
Log likelihood =  -2022.198                     Pseudo R2         =     0.1455

--------------------------------------------------------------------------------
               |   Observed   Bootstrap                         Normal-based
      partic5a |        RRR   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
Part_in_rev    |
        gender |   1.494754   .2372636     2.53   0.011      1.09511    2.040242
        newage |   .9723597   .0049189    -5.54   0.000     .9627666    .9820485
      edulevel |   1.166069   .0791269     2.26   0.024     1.020854    1.331941
 consumergoods |   1.216015   .0434248     5.48   0.000     1.133814    1.304175
ukrspeakathome |   3.137825   .5338268     6.72   0.000     2.248107    4.379662
         _cons |   .1432741   .0608064    -4.58   0.000      .062361    .3291718
---------------+----------------------------------------------------------------
Aid_rev        |
        gender |   1.335445   .3707684     1.04   0.297     .7749993     2.30118
        newage |   .9974216   .0095836    -0.27   0.788     .9788138    1.016383
      edulevel |    1.02673   .1368136     0.20   0.843     .7907368    1.333154
 consumergoods |   1.130487   .0713146     1.94   0.052     .9990088     1.27927
ukrspeakathome |    2.38496   .7662417     2.71   0.007     1.270593    4.476677
         _cons |   .0257419   .0215167    -4.38   0.000     .0050021    .1324738
---------------+----------------------------------------------------------------
Support_rev    |  (base outcome)
---------------+----------------------------------------------------------------
Apathetic      |
        gender |   .7300962   .1368598    -1.68   0.093     .5056134    1.054245
        newage |   .9727284    .006482    -4.15   0.000     .9601065    .9855162
      edulevel |   1.001542   .0842092     0.02   0.985     .8493767    1.180967
 consumergoods |   .9202741   .0409322    -1.87   0.062     .8434458      1.0041
ukrspeakathome |   .2301635   .0486723    -6.95   0.000     .1520667    .3483684
         _cons |   2.407614   1.006006     2.10   0.035     1.061495    5.460798
---------------+----------------------------------------------------------------
Oppose_rev     |
        gender |   .8689785   .1144416    -1.07   0.286     .6712877    1.124888
        newage |   .9985631   .0040138    -0.36   0.721      .990727    1.006461
      edulevel |   .8959331   .0544642    -1.81   0.071     .7952994    1.009301
 consumergoods |   1.012041   .0282322     0.43   0.668     .9581924    1.068916
ukrspeakathome |   .0887285   .0136978   -15.69   0.000     .0655625      .12008
         _cons |   2.714424   .7899224     3.43   0.001     1.534512    4.801591
--------------------------------------------------------------------------------
Note: One or more parameters could not be estimated in 80 bootstrap replicates;
      standard-error estimates include only complete replications.

. estat bootstrap, all

Multinomial logistic regression                 Number of obs     =      1,746
                                                Replications      =        920

------------------------------------------------------------------------------
             |    Observed               Bootstrap
    partic5a |       Coef.       Bias    Std. Err.  [95% Conf. Interval]
-------------+----------------------------------------------------------------
Part_in_rev  |
      gender |   .40196167  -.0108264   .15873086    .0908549   .7130685   (N)
             |                                       .0829748   .6942806   (P)
             |                                       .1093158   .7322342  (BC)
             |                                       .1093158   .7319939 (BCa)
      newage |  -.02802943  -.0003857    .0050587   -.0379443  -.0181146   (N)
             |                                       -.038454  -.0177953   (P)
             |                                      -.0375099  -.0172175  (BC)
             |                                      -.0375099    -.01717 (BCa)
    edulevel |   .15363816   .0006272   .06785781    .0206393    .286637   (N)
             |                                       .0194368   .2858852   (P)
             |                                       .0188591    .284481  (BC)
             |                                       .0188591    .284481 (BCa)
consumergo~s |   .19557888   .0018248   .03571077    .1255871   .2655707   (N)
             |                                        .131062    .268256   (P)
             |                                        .129652   .2653485  (BC)
             |                                       .1291154   .2653485 (BCa)
ukrspeakat~e |     1.14353   .0101353   .17012636    .8100885   1.476972   (N)
             |                                       .8424817   1.489121   (P)
             |                                       .8348711    1.48033  (BC)
             |                                       .8305748   1.478559 (BCa)
       _cons |  -1.9429956  -.0109926   .42440578   -2.774816  -1.111176   (N)
             |                                      -2.775203  -1.116468   (P)
             |                                      -2.727793  -1.097025  (BC)
             |                                      -2.718888  -1.092416 (BCa)
-------------+----------------------------------------------------------------
Aid_rev      |
      gender |   .28926451  -.0038331   .27763659   -.2548932   .8334222   (N)
             |                                      -.2420941   .8843529   (P)
             |                                      -.2278477   .8958694  (BC)
             |                                      -.2278477   .8958694 (BCa)
      newage |  -.00258171   -.000071   .00960839   -.0214138   .0162504   (N)
             |                                      -.0214467   .0164111   (P)
             |                                      -.0203486   .0172811  (BC)
             |                                      -.0208995   .0172127 (BCa)
    edulevel |   .02637869  -.0044835   .13325186   -.2347902   .2875475   (N)
             |                                      -.2478753   .2648178   (P)
             |                                      -.2464911   .2649392  (BC)
             |                                      -.2251712   .2711249 (BCa)
consumergo~s |   .12264874   .0039097   .06308304   -.0009917   .2462892   (N)
             |                                       .0009324   .2532486   (P)
             |                                      -.0229465     .23848  (BC)
             |                                       -.023226   .2383105 (BCa)
ukrspeakat~e |   .86918233   .0310086   .32128074    .2394837   1.498881   (N)
             |                                       .2979958     1.6081   (P)
             |                                       .2803587   1.570032  (BC)
             |                                       .2415448   1.524813 (BCa)
       _cons |  -3.6596361  -.0892267   .83586521   -5.297902   -2.02137   (N)
             |                                      -5.442281  -2.215934   (P)
             |                                      -5.261242  -2.104564  (BC)
             |                                      -5.236734  -2.058749 (BCa)
-------------+----------------------------------------------------------------
Support_rev  |
    o.gender |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
    o.newage |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
  o.edulevel |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
o.consumer~s |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
o.ukrspeak~e |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
     o._cons |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
-------------+----------------------------------------------------------------
Apathetic    |
      gender |  -.31457899   -.001022   .18745446    -.681983    .052825   (N)
             |                                      -.7177691   .0172729   (P)
             |                                      -.7281956   .0108285  (BC)
             |                                      -.7249762   .0118898 (BCa)
      newage |  -.02765036  -.0002304   .00666374   -.0407111  -.0145897   (N)
             |                                      -.0402094   -.015177   (P)
             |                                      -.0400176  -.0149664  (BC)
             |                                      -.0399795  -.0148605 (BCa)
    edulevel |   .00154049  -.0003915   .08407957   -.1632524   .1663334   (N)
             |                                       -.176421   .1649018   (P)
             |                                       -.179889   .1576923  (BC)
             |                                      -.1778354   .1586361 (BCa)
consumergo~s |  -.08308375    .000201   .04447829   -.1702596   .0040921   (N)
             |                                      -.1701275    .005405   (P)
             |                                      -.1746446   .0037515  (BC)
             |                                       -.172544   .0038579 (BCa)
ukrspeakat~e |  -1.4689653  -.0085648   .21146841   -1.883436  -1.054495   (N)
             |                                      -1.887025  -1.055548   (P)
             |                                      -1.884171  -1.055174  (BC)
             |                                      -1.874396  -1.048148 (BCa)
       _cons |    .8786364  -.0130011   .41784364    .0596779   1.697595   (N)
             |                                       .0306628   1.664669   (P)
             |                                       .0908428   1.685249  (BC)
             |                                       .0908428   1.685249 (BCa)
-------------+----------------------------------------------------------------
Oppose_rev   |
      gender |  -.14043684  -.0080709   .13169664   -.3985575   .1176838   (N)
             |                                      -.4085007   .1095344   (P)
             |                                      -.3983273   .1205014  (BC)
             |                                      -.3983273   .1205014 (BCa)
      newage |  -.00143798  -.0000463   .00401962   -.0093163   .0064403   (N)
             |                                      -.0095958   .0066557   (P)
             |                                      -.0096395    .006643  (BC)
             |                                      -.0096395    .006643 (BCa)
    edulevel |  -.10988952  -.0011787   .06079047   -.2290367   .0092576   (N)
             |                                      -.2327182   .0065567   (P)
             |                                      -.2304654   .0135791  (BC)
             |                                      -.2304654   .0135791 (BCa)
consumergo~s |   .01196906   .0006686   .02789631   -.0427067   .0666448   (N)
             |                                      -.0423976   .0684877   (P)
             |                                      -.0422868   .0691159  (BC)
             |                                      -.0422868   .0691159 (BCa)
ukrspeakat~e |  -2.4221744  -.0066063   .15437919   -2.724752  -2.119597   (N)
             |                                      -2.733954  -2.135409   (P)
             |                                      -2.731022  -2.131747  (BC)
             |                                      -2.726097  -2.128074 (BCa)
       _cons |   .99857992  -.0073422   .29100916    .4282124   1.568947   (N)
             |                                        .447303   1.575896   (P)
             |                                       .4616867   1.608318  (BC)
             |                                       .4616867   1.608318 (BCa)
------------------------------------------------------------------------------
(N)    normal confidence interval
(P)    percentile confidence interval
(BC)   bias-corrected confidence interval
(BCa)  bias-corrected and accelerated confidence interval
Note: One or more parameters could not be estimated in 80 bootstrap replicates;
      standard-error estimates include only complete replications.

. *       RESULT:  No changes
. * Testing for multicollinearity
. collin  gender newage edulevel consumergoods ukrspeakathome if EVA_vers=="yr2005"
(obs=1,799)

  Collinearity Diagnostics

                        SQRT                   R-
  Variable      VIF     VIF    Tolerance    Squared
----------------------------------------------------
    gender      1.02    1.01    0.9793      0.0207
    newage      1.18    1.09    0.8449      0.1551
  edulevel      1.22    1.10    0.8220      0.1780
consumergoods      1.12    1.06    0.8956      0.1044
ukrspeakathome      1.02    1.01    0.9766      0.0234
----------------------------------------------------
  Mean VIF      1.11

                           Cond
        Eigenval          Index
---------------------------------
    1     4.5314          1.0000
    2     0.6090          2.7278
    3     0.4629          3.1288
    4     0.2309          4.4303
    5     0.1362          5.7676
    6     0.0297         12.3614
---------------------------------
 Condition Number        12.3614 
 Eigenvalues & Cond Index computed from scaled raw sscp (w/ intercept)
 Det(correlation matrix)    0.7534

. *       RESULT:  tolerances are all > .5
. 
. * Euromaidan Revolution
. * checkrob command
. * checkrob 2 3 ch7fig7_9b.txt: mlogit partic5a gender newage edulevel consumergoods ukrspeakathome if version==1
> , rrr
. *       Reworked results can be found in the Excel file checkrob.results.chapter7.xlsx,
. *               under the Robustnesstests/Outputfiles directory
. *       RESULT:  All patterns of significance stable (with exception of edulevel, which is marginally significan
> t at the .10 level when consumergoods included but ukrspeakathome not included); all signs stable
. * Robust standard errors
. mlogit partic5a  gender newage edulevel consumergoods ukrspeakathome  if EVA_vers=="yr2014", rrr vce(robust)

Iteration 0:   log pseudolikelihood = -2414.9662  
Iteration 1:   log pseudolikelihood =  -2223.672  
Iteration 2:   log pseudolikelihood = -2213.3862  
Iteration 3:   log pseudolikelihood = -2213.3155  
Iteration 4:   log pseudolikelihood = -2213.3155  

Multinomial logistic regression                 Number of obs     =      1,736
                                                Wald chi2(20)     =     300.72
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -2213.3155               Pseudo R2         =     0.0835

--------------------------------------------------------------------------------
               |               Robust
      partic5a |        RRR   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
Part_in_rev    |
        gender |   1.588197     .30218     2.43   0.015     1.093829    2.305999
        newage |   .9859188   .0059749    -2.34   0.019     .9742775    .9976992
      edulevel |   1.175512   .0919963     2.07   0.039      1.00835    1.370385
 consumergoods |   1.140657   .0487029     3.08   0.002     1.049086     1.24022
ukrspeakathome |   2.829424   .5987728     4.91   0.000     1.868803    4.283831
         _cons |   .0425023   .0228782    -5.87   0.000     .0147988    .1220671
---------------+----------------------------------------------------------------
Aid_rev        |
        gender |    1.00039   .1932923     0.00   0.998     .6850189    1.460951
        newage |   .9961001   .0060223    -0.65   0.518     .9843662    1.007974
      edulevel |   1.241997   .1005497     2.68   0.007     1.059763    1.455567
 consumergoods |   1.179995   .0506818     3.85   0.000     1.084727     1.28363
ukrspeakathome |   2.612066   .5625329     4.46   0.000     1.712655    3.983808
         _cons |   .0236864   .0124005    -7.15   0.000     .0084893     .066089
---------------+----------------------------------------------------------------
Support_rev    |  (base outcome)
---------------+----------------------------------------------------------------
Apathetic      |
        gender |   .6359465   .1024448    -2.81   0.005     .4637674     .872049
        newage |    .994295   .0048947    -1.16   0.245     .9847477    1.003935
      edulevel |   .9912426   .0668353    -0.13   0.896     .8685344    1.131287
 consumergoods |   .8444038   .0348567    -4.10   0.000     .7787764    .9155615
ukrspeakathome |   .3980307   .0659659    -5.56   0.000     .2876381     .550791
         _cons |   1.746346   .7105631     1.37   0.171     .7866627    3.876787
---------------+----------------------------------------------------------------
Oppose_rev     |
        gender |   .8542415   .1064877    -1.26   0.206     .6690706     1.09066
        newage |   1.001084    .003877     0.28   0.780     .9935137    1.008712
      edulevel |   .9822968   .0530486    -0.33   0.741      .883637    1.091972
 consumergoods |    .894853   .0249942    -3.98   0.000      .847182    .9452065
ukrspeakathome |   .1712323   .0243831   -12.39   0.000     .1295319    .2263575
         _cons |   2.503317   .8113615     2.83   0.005     1.326261    4.725009
--------------------------------------------------------------------------------

. *       RESULT:  No changes
. * Bootstrapped standard errors
. mlogit partic5a  gender newage edulevel consumergoods ukrspeakathome if EVA_vers=="yr2014", rrr vce(bootstrap, b
> ca seed(1234) rep(1000))
(running mlogit on estimation sample)

Jackknife replications (1736)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400
..................................................   450
..................................................   500
..................................................   550
..................................................   600
..................................................   650
..................................................   700
..................................................   750
..................................................   800
..................................................   850
..................................................   900
..................................................   950
..................................................  1000
..................................................  1050
..................................................  1100
..................................................  1150
..................................................  1200
..................................................  1250
..................................................  1300
..................................................  1350
..................................................  1400
..................................................  1450
..................................................  1500
..................................................  1550
..................................................  1600
..................................................  1650
..................................................  1700
....................................

Bootstrap replications (1000)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400
..................................................   450
..................................................   500
..................................................   550
..................................................   600
..................................................   650
..................................................   700
..................................................   750
..................................................   800
..................................................   850
..................................................   900
..................................................   950
..................................................  1000

Multinomial logistic regression                 Number of obs     =      1,736
                                                Replications      =      1,000
                                                Wald chi2(20)     =     301.14
                                                Prob > chi2       =     0.0000
Log likelihood = -2213.3155                     Pseudo R2         =     0.0835

--------------------------------------------------------------------------------
               |   Observed   Bootstrap                         Normal-based
      partic5a |        RRR   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
Part_in_rev    |
        gender |   1.588197   .3171116     2.32   0.021     1.073858    2.348885
        newage |   .9859188   .0058116    -2.41   0.016     .9745938    .9973753
      edulevel |   1.175512   .0930561     2.04   0.041      1.00657    1.372808
 consumergoods |   1.140657   .0500676     3.00   0.003     1.046628    1.243132
ukrspeakathome |   2.829424   .6078946     4.84   0.000     1.857032    4.310985
         _cons |   .0425023   .0227894    -5.89   0.000     .0148595    .1215685
---------------+----------------------------------------------------------------
Aid_rev        |
        gender |    1.00039   .2013812     0.00   0.998     .6742484    1.484289
        newage |   .9961001   .0061222    -0.64   0.525     .9841729    1.008172
      edulevel |   1.241997   .1036407     2.60   0.009     1.054607    1.462684
 consumergoods |   1.179995   .0508398     3.84   0.000     1.084442    1.283967
ukrspeakathome |   2.612066   .5720128     4.38   0.000     1.700516    4.012247
         _cons |   .0236864   .0121932    -7.27   0.000     .0086362    .0649649
---------------+----------------------------------------------------------------
Support_rev    |  (base outcome)
---------------+----------------------------------------------------------------
Apathetic      |
        gender |   .6359465    .100385    -2.87   0.004      .466721    .8665304
        newage |    .994295   .0048329    -1.18   0.239     .9848677    1.003813
      edulevel |   .9912426   .0639909    -0.14   0.892     .8734329    1.124943
 consumergoods |   .8444038   .0346478    -4.12   0.000     .7791543    .9151175
ukrspeakathome |   .3980307   .0671652    -5.46   0.000     .2859445    .5540533
         _cons |   1.746346   .7042639     1.38   0.167     .7922439    3.849476
---------------+----------------------------------------------------------------
Oppose_rev     |
        gender |   .8542415   .1075264    -1.25   0.211      .667478    1.093262
        newage |   1.001084   .0039016     0.28   0.781     .9934659     1.00876
      edulevel |   .9822968   .0542244    -0.32   0.746     .8815663    1.094537
 consumergoods |    .894853    .025641    -3.88   0.000     .8459827    .9465465
ukrspeakathome |   .1712323   .0246767   -12.25   0.000     .1290973    .2271194
         _cons |   2.503317   .8097065     2.84   0.005      1.32798    4.718891
--------------------------------------------------------------------------------

. estat bootstrap, all

Multinomial logistic regression                 Number of obs     =      1,736
                                                Replications      =       1000

------------------------------------------------------------------------------
             |    Observed               Bootstrap
    partic5a |       Coef.       Bias    Std. Err.  [95% Conf. Interval]
-------------+----------------------------------------------------------------
Part_in_rev  |
      gender |   .46259926   .0125483   .19966767    .0712578   .8539407   (N)
             |                                       .0624163    .853881   (P)
             |                                       .0432137   .8324343  (BC)
             |                                       .0432137   .8300015 (BCa)
      newage |  -.01418128    -.00005   .00589459   -.0257345  -.0026281   (N)
             |                                      -.0258457  -.0036141   (P)
             |                                       -.025714  -.0035033  (BC)
             |                                       -.025714  -.0035033 (BCa)
    edulevel |   .16170352   .0022973   .07916218    .0065485   .3168585   (N)
             |                                       .0051737   .3225724   (P)
             |                                         .00162   .3192818  (BC)
             |                                       .0026459   .3198164 (BCa)
consumergo~s |   .13160401   .0002987   .04389371    .0455739   .2176341   (N)
             |                                       .0527393   .2159378   (P)
             |                                       .0535155   .2164992  (BC)
             |                                       .0529947   .2164992 (BCa)
ukrspeakat~e |    1.040073   .0248709   .21484752    .6189796   1.461166   (N)
             |                                       .6447136   1.505183   (P)
             |                                       .5806105   1.431146  (BC)
             |                                       .5660886   1.422189 (BCa)
       _cons |  -3.1581977   -.045318   .53619348   -4.209118  -2.107278   (N)
             |                                      -4.271231  -2.172226   (P)
             |                                       -4.21645  -2.149698  (BC)
             |                                      -4.214783  -2.144964 (BCa)
-------------+----------------------------------------------------------------
Aid_rev      |
      gender |   .00038949   .0008273   .20130278   -.3941567   .3949357   (N)
             |                                      -.3694107   .4077576   (P)
             |                                      -.3799089   .4019543  (BC)
             |                                      -.3799089   .4042594 (BCa)
      newage |  -.00390748  -.0002555   .00614613   -.0159537   .0081387   (N)
             |                                      -.0164449    .008201   (P)
             |                                      -.0153317   .0100286  (BC)
             |                                      -.0153317   .0100286 (BCa)
    edulevel |    .2167206  -.0015873   .08344681    .0531679   .3802733   (N)
             |                                       .0442958   .3832345   (P)
             |                                       .0488926   .3862936  (BC)
             |                                       .0496201   .3862936 (BCa)
consumergo~s |   .16551015   .0030271   .04308473    .0810656   .2499547   (N)
             |                                       .0847318   .2569325   (P)
             |                                       .0824684   .2469933  (BC)
             |                                       .0824684   .2469933 (BCa)
ukrspeakat~e |   .96014151   .0121115   .21898863    .5309317   1.389351   (N)
             |                                       .5535776   1.408099   (P)
             |                                       .5443037   1.395118  (BC)
             |                                       .5321507   1.381844 (BCa)
       _cons |  -3.7428536  -.0319135   .51477718   -4.751798  -2.733909   (N)
             |                                       -4.85567  -2.787609   (P)
             |                                      -4.780946  -2.748298  (BC)
             |                                      -4.761055  -2.741876 (BCa)
-------------+----------------------------------------------------------------
Support_rev  |
    o.gender |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
    o.newage |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
  o.edulevel |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
o.consumer~s |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
o.ukrspeak~e |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
     o._cons |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
-------------+----------------------------------------------------------------
Apathetic    |
      gender |  -.45264087   .0072347   .15785127   -.7620237  -.1432581   (N)
             |                                       -.729678  -.1449805   (P)
             |                                      -.7854658   -.174944  (BC)
             |                                      -.7728905  -.1732875 (BCa)
      newage |  -.00572134   -.000062   .00486062    -.015248   .0038053   (N)
             |                                      -.0152027   .0036417   (P)
             |                                      -.0152167   .0036385  (BC)
             |                                      -.0152167   .0036385 (BCa)
    edulevel |  -.00879601  -.0026672   .06455629    -.135324    .117732   (N)
             |                                      -.1367725   .1146548   (P)
             |                                      -.1368092   .1144041  (BC)
             |                                      -.1367359   .1149055 (BCa)
consumergo~s |  -.16912452  -.0001618   .04103223   -.2495462  -.0887028   (N)
             |                                      -.2508811  -.0922715   (P)
             |                                      -.2505625  -.0921239  (BC)
             |                                      -.2492754  -.0905275 (BCa)
ukrspeakat~e |  -.92122603   .0014623   .16874372   -1.251958  -.5904944   (N)
             |                                       -1.24992  -.5544633   (P)
             |                                       -1.26476  -.5661688  (BC)
             |                                      -1.253972  -.5578416 (BCa)
       _cons |   .55752555   .0024525   .40327859    -.232886   1.347937   (N)
             |                                      -.2549542   1.360713   (P)
             |                                      -.2452652   1.360896  (BC)
             |                                      -.2452652   1.360896 (BCa)
-------------+----------------------------------------------------------------
Oppose_rev   |
      gender |   -.1575413  -.0001306   .12587353   -.4042489   .0891663   (N)
             |                                      -.3962265     .08548   (P)
             |                                      -.3958696   .0856911  (BC)
             |                                      -.3958696   .0856911 (BCa)
      newage |   .00108319  -.0000949   .00389741   -.0065556    .008722   (N)
             |                                      -.0067625   .0088915   (P)
             |                                      -.0064505   .0093349  (BC)
             |                                      -.0064505   .0093349 (BCa)
    edulevel |  -.01786181  -.0016848   .05520167   -.1260551   .0903315   (N)
             |                                      -.1286494   .0923547   (P)
             |                                       -.127054   .0965186  (BC)
             |                                      -.1264877   .0965186 (BCa)
consumergo~s |  -.11109579  -.0008672   .02865389   -.1672564  -.0549352   (N)
             |                                      -.1665946  -.0544569   (P)
             |                                      -.1647217  -.0526514  (BC)
             |                                      -.1647217  -.0524477 (BCa)
ukrspeakat~e |  -1.7647339  -.0058063   .14411218   -2.047189  -1.482279   (N)
             |                                      -2.080943  -1.496366   (P)
             |                                      -2.089967  -1.502187  (BC)
             |                                      -2.080948  -1.496449 (BCa)
       _cons |   .91761649   .0148683   .32345351    .2836593   1.551574   (N)
             |                                       .3169215   1.552799   (P)
             |                                       .2429066   1.491716  (BC)
             |                                       .2429066   1.491716 (BCa)
------------------------------------------------------------------------------
(N)    normal confidence interval
(P)    percentile confidence interval
(BC)   bias-corrected confidence interval
(BCa)  bias-corrected and accelerated confidence interval

. *       RESULT:  No changes
. * Testing for multicollinearity
. collin  gender newage edulevel consumergoods ukrspeakathome if EVA_vers=="yr2014"
(obs=1,783)

  Collinearity Diagnostics

                        SQRT                   R-
  Variable      VIF     VIF    Tolerance    Squared
----------------------------------------------------
    gender      1.02    1.01    0.9772      0.0228
    newage      1.07    1.03    0.9369      0.0631
  edulevel      1.06    1.03    0.9453      0.0547
consumergoods      1.06    1.03    0.9390      0.0610
ukrspeakathome      1.02    1.01    0.9832      0.0168
----------------------------------------------------
  Mean VIF      1.05

                           Cond
        Eigenval          Index
---------------------------------
    1     4.6469          1.0000
    2     0.5497          2.9075
    3     0.4988          3.0522
    4     0.1585          5.4153
    5     0.1189          6.2504
    6     0.0272         13.0707
---------------------------------
 Condition Number        13.0707 
 Eigenvalues & Cond Index computed from scaled raw sscp (w/ intercept)
 Det(correlation matrix)    0.8898

. *       RESULT:  tolerances are all > .5
. 
. 
. * =================================================================
. * ROBUSTNESS TESTS FOR FIGURE 7.10: RELIGIOUS PRACTICE IN TUNISIAN 
. *               AND EGYPTIANS REVOLUTIONS
. * =================================================================
. clear

. use fullarabbarom2.dta

. * Egyptian Revolution
. * checkrob command
. * checkrob 1 3 ch7fig7_10a.txt: mlogit egpartic5 gender newage religscale christian if country==2 [pw=WT], rrr b
> (3)
. *       Reworked results can be found in the Excel file checkrob.results.chapter7.xlsx,
. *               under the Robustnesstests/Outputfiles directory
. *       RESULT:  All patterns of significance stable (with exception of religscale, which is significant at the 
> .10 level when newage is excluded); all signs stable; christian is statistically insignificant
. * Robust standard errors
. mlogit egpartic5 gender newage  religscale christian if country==2 [pw=WT], rrr b(3) vce(robust)

Iteration 0:   log pseudolikelihood = -1206.3169  
Iteration 1:   log pseudolikelihood = -1164.3115  
Iteration 2:   log pseudolikelihood =  -1160.798  
Iteration 3:   log pseudolikelihood = -1160.6491  
Iteration 4:   log pseudolikelihood = -1160.6194  
Iteration 5:   log pseudolikelihood =  -1160.613  
Iteration 6:   log pseudolikelihood = -1160.6116  
Iteration 7:   log pseudolikelihood = -1160.6112  
Iteration 8:   log pseudolikelihood = -1160.6112  
Iteration 9:   log pseudolikelihood = -1160.6111  

Multinomial logistic regression                 Number of obs     =      1,207
                                                Wald chi2(16)     =    3240.62
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -1160.6111               Pseudo R2         =     0.0379

------------------------------------------------------------------------------
             |               Robust
   egpartic5 |        RRR   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      gender |   3.269492   .8559003     4.53   0.000     1.957264    5.461489
      newage |   .9805408   .0077838    -2.48   0.013     .9654029    .9959161
  religscale |   1.085508   .0440135     2.02   0.043     1.002582    1.175293
   christian |   .7347861   .4658889    -0.49   0.627     .2120602    2.546025
       _cons |    .051202   .0246313    -6.18   0.000     .0199437    .1314522
-------------+----------------------------------------------------------------
Aid          |
      gender |   1.172399   .4834881     0.39   0.700     .5224525    2.630899
      newage |   1.004481   .0122646     0.37   0.714     .9807279    1.028809
  religscale |   1.163786   .0717756     2.46   0.014     1.031278    1.313319
   christian |   1.19e-06   3.27e-07   -49.81   0.000     6.98e-07    2.04e-06
       _cons |   .0066187   .0048953    -6.78   0.000     .0015532    .0282052
-------------+----------------------------------------------------------------
Support      |  (base outcome)
-------------+----------------------------------------------------------------
Apathetic    |
      gender |   1.027114   .2449547     0.11   0.911     .6436013    1.639156
      newage |   1.014772   .0087056     1.71   0.087      .997852    1.031979
  religscale |   .8819621   .0441585    -2.51   0.012     .7995241    .9729001
   christian |   2.896803   1.056756     2.92   0.004     1.417098    5.921584
       _cons |   .1326261   .0617442    -4.34   0.000     .0532539    .3302984
-------------+----------------------------------------------------------------
Oppose       |
      gender |   .6653437   .1156928    -2.34   0.019     .4731899    .9355276
      newage |   .9960126   .0063974    -0.62   0.534     .9835525    1.008631
  religscale |   .9030543   .0265095    -3.47   0.001      .852563    .9565358
   christian |   1.938829   .6144969     2.09   0.037     1.041736    3.608455
       _cons |   .6570219   .2305478    -1.20   0.231     .3302895    1.306968
------------------------------------------------------------------------------

. *       RESULT:  No changes
. * Bootstrapped standard errors
. mlogit egpartic5 gender newage  religscale christian if country==2, rrr b(3) vce(bootstrap, bca seed(1234) rep(1
> 000))
(running mlogit on estimation sample)

Jackknife replications (1207)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400
..................................................   450
..................................................   500
..................................................   550
..................................................   600
..................................................   650
..................................................   700
..................................................   750
..................................................   800
..................................................   850
..................................................   900
..................................................   950
..................................................  1000
..................................................  1050
..................................................  1100
..................................................  1150
..................................................  1200
.......

Bootstrap replications (1000)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400
..................................................   450
..................................................   500
..................................................   550
..................................................   600
..................................................   650
..................................................   700
..................................................   750
..................................................   800
..................................................   850
..................................................   900
..................................................   950
..................................................  1000

Multinomial logistic regression                 Number of obs     =      1,207
                                                Replications      =      1,000
                                                Wald chi2(16)     =     616.59
                                                Prob > chi2       =     0.0000
Log likelihood =  -1163.597                     Pseudo R2         =     0.0372

------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
   egpartic5 |        RRR   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      gender |   3.276317      .8761     4.44   0.000     1.939864    5.533508
      newage |   .9799497   .0076618    -2.59   0.010     .9650473    .9950821
  religscale |   1.101639   .0452448     2.36   0.018     1.016436    1.193984
   christian |   .7548841    2.43627    -0.09   0.931     .0013513    421.6959
       _cons |   .0471507   .0224364    -6.42   0.000     .0185544    .1198196
-------------+----------------------------------------------------------------
Aid          |
      gender |   1.063264   .4578707     0.14   0.887     .4571834    2.472817
      newage |   1.006594   .0122719     0.54   0.590      .982827    1.030937
  religscale |   1.151146   .0692271     2.34   0.019     1.023155    1.295149
   christian |   1.20e-06   8.00e-07   -20.44   0.000     3.25e-07    4.44e-06
       _cons |   .0071774   .0053375    -6.64   0.000      .001671    .0308292
-------------+----------------------------------------------------------------
Support      |  (base outcome)
-------------+----------------------------------------------------------------
Apathetic    |
      gender |   1.052822   .2568263     0.21   0.833     .6526979    1.698235
      newage |   1.016039    .008321     1.94   0.052     .9998605     1.03248
  religscale |   .8822678   .0424502    -2.60   0.009     .8028694    .9695181
   christian |   2.756682   1.033663     2.70   0.007     1.321949    5.748554
       _cons |   .1267989   .0590042    -4.44   0.000     .0509354    .3156539
-------------+----------------------------------------------------------------
Oppose       |
      gender |   .7035996   .1231032    -2.01   0.045     .4993415    .9914104
      newage |   .9960031    .006374    -0.63   0.531     .9835884    1.008574
  religscale |    .909012   .0269079    -3.22   0.001     .8577741    .9633105
   christian |   1.909991   .6281665     1.97   0.049       1.0025    3.638969
       _cons |    .603558   .2059028    -1.48   0.139     .3092666    1.177891
------------------------------------------------------------------------------

. estat bootstrap, all

Multinomial logistic regression                 Number of obs     =      1,207
                                                Replications      =       1000

------------------------------------------------------------------------------
             |    Observed               Bootstrap
   egpartic5 |       Coef.       Bias    Std. Err.  [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      gender |     1.18672   .0211679   .26740392    .6626179   1.710822   (N)
             |                                       .7304776   1.770466   (P)
             |                                       .7197775    1.74563  (BC)
             |                                       .7117531   1.718125 (BCa)
      newage |  -.02025407  -.0001948   .00781854   -.0355781    -.00493   (N)
             |                                      -.0350711  -.0052077   (P)
             |                                      -.0345789  -.0042542  (BC)
             |                                       -.034352   -.004134 (BCa)
  religscale |   .09679888   .0032403   .04107044    .0163023   .1772955   (N)
             |                                       .0195334   .1769414   (P)
             |                                       .0169438   .1738851  (BC)
             |                                       .0166836   .1722409 (BCa)
   christian |  -.28119101  -.8648789   3.2273428   -6.606667   6.044285   (N)
             |                                      -14.09426   .6535716   (P)
             |                                        -13.778   .8023461  (BC)
             |                                      -1.652397   .9660931 (BCa)
       _cons |  -3.0544074  -.0557322   .47584535   -3.987047  -2.121768   (N)
             |                                      -4.096598  -2.223011   (P)
             |                                      -3.944181  -2.092823  (BC)
             |                                      -3.906196  -2.084509 (BCa)
-------------+----------------------------------------------------------------
Aid          |
      gender |   .06134364  -.0386573   .43062742   -.7826706   .9053579   (N)
             |                                       -.800024   .8636607   (P)
             |                                      -.7358775   .9591364  (BC)
             |                                       -.735907   .9533877 (BCa)
      newage |   .00657276   .0008597   .01219151   -.0173221   .0304677   (N)
             |                                       -.016854   .0302402   (P)
             |                                      -.0183141   .0291623  (BC)
             |                                      -.0183141   .0291623 (BCa)
  religscale |     .140758   .0027908   .06013753    .0228906   .2586254   (N)
             |                                       .0302379   .2617436   (P)
             |                                        .027765    .260822  (BC)
             |                                       .0313615   .2620792 (BCa)
   christian |  -13.633133   .4463811     .666943   -14.94032  -12.32595   (N)
             |                                      -14.38838  -12.10826   (P)
             |                                      -15.94105  -12.56048  (BC)
             |                                      -17.63887  -12.59044 (BCa)
       _cons |  -4.9368247  -.1037028   .74365248   -6.394357  -3.479293   (N)
             |                                      -6.692125   -3.66992   (P)
             |                                      -6.548615  -3.580139  (BC)
             |                                      -6.462977  -3.519711 (BCa)
-------------+----------------------------------------------------------------
Support      |
    o.gender |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
    o.newage |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
o.religscale |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
 o.christian |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
     o._cons |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
-------------+----------------------------------------------------------------
Apathetic    |
      gender |   .05147427   .0046259   .24394081   -.4266409   .5295895   (N)
             |                                      -.4146469   .5349025   (P)
             |                                      -.4314664    .515938  (BC)
             |                                      -.4314664    .515938 (BCa)
      newage |   .01591191   .0003217   .00818965   -.0001395   .0319633   (N)
             |                                      -.0003683   .0321908   (P)
             |                                      -.0012595   .0311173  (BC)
             |                                      -.0012595   .0311173 (BCa)
  religscale |  -.12525969  -.0028919   .04811493   -.2195632  -.0309562   (N)
             |                                       -.220962   -.030622   (P)
             |                                      -.2191221  -.0281439  (BC)
             |                                      -.2182673  -.0256828 (BCa)
   christian |   1.0140277  -.0257823   .37496639    .2791071   1.748948   (N)
             |                                       .1771447   1.673011   (P)
             |                                       .1617188   1.651573  (BC)
             |                                       .1875377   1.688572 (BCa)
       _cons |   -2.065153  -.0219397   .46533711   -2.977197  -1.153109   (N)
             |                                      -3.043877   -1.17669   (P)
             |                                      -3.010498  -1.115439  (BC)
             |                                      -2.988989  -1.086951 (BCa)
-------------+----------------------------------------------------------------
Oppose       |
      gender |  -.35154589   .0015282   .17496198   -.6944651  -.0086267   (N)
             |                                      -.7005061  -.0094338   (P)
             |                                      -.6985246    -.00721  (BC)
             |                                      -.6971155   .0143699 (BCa)
      newage |  -.00400491  -.0002161   .00639953   -.0165478   .0085379   (N)
             |                                       -.016892   .0075432   (P)
             |                                      -.0167101   .0075745  (BC)
             |                                      -.0166665   .0076938 (BCa)
  religscale |  -.09539698  -.0007129   .02960131   -.1534145  -.0373795   (N)
             |                                      -.1517445  -.0371904   (P)
             |                                       -.148869  -.0351706  (BC)
             |                                      -.1486836  -.0341654 (BCa)
   christian |   .64709877  -.0170121   .32888446    .0024971     1.2917   (N)
             |                                      -.0343794   1.231759   (P)
             |                                      -.0027464   1.258675  (BC)
             |                                       .0249073    1.32194 (BCa)
       _cons |  -.50491312   .0058456   .34114835   -1.173552   .1637254   (N)
             |                                      -1.151367   .1920935   (P)
             |                                      -1.152142    .183278  (BC)
             |                                      -1.150591    .200909 (BCa)
------------------------------------------------------------------------------
(N)    normal confidence interval
(P)    percentile confidence interval
(BC)   bias-corrected confidence interval
(BCa)  bias-corrected and accelerated confidence interval

. *       RESULT:  No changes
. * Testing for multicollinearity
. collin gender newage  religscale christian if country==2
(obs=1,215)

  Collinearity Diagnostics

                        SQRT                   R-
  Variable      VIF     VIF    Tolerance    Squared
----------------------------------------------------
    gender      1.02    1.01    0.9821      0.0179
    newage      1.03    1.01    0.9737      0.0263
religscale      1.02    1.01    0.9791      0.0209
 christian      1.00    1.00    0.9956      0.0044
----------------------------------------------------
  Mean VIF      1.02

                           Cond
        Eigenval          Index
---------------------------------
    1     3.5283          1.0000
    2     0.9281          1.9497
    3     0.4043          2.9542
    4     0.0967          6.0414
    5     0.0426          9.1025
---------------------------------
 Condition Number         9.1025 
 Eigenvalues & Cond Index computed from scaled raw sscp (w/ intercept)
 Det(correlation matrix)    0.9648

. *       RESULT:  tolerances are all > .5
. * Tunisian Revolution
. * checkrob command
. * checkrob 1 2 ch7fig7_10b.txt: mlogit tpartic4 gender newage religscale if country==10 [pw=WT], rrr b(2)
. *       Reworked results can be found in the Excel file checkrob.results.chapter7.xlsx,
. *               under the Robustnesstests/Outputfiles directory
. *       RESULT:  All patterns of significance stable (with exception of religscale, which is insignificant when 
> newage is excluded); all signs stable
. * Robust standard errors
. mlogit tpartic4 gender newage religscale if country==10 [pw=WT], rrr b(2) vce(robust)

Iteration 0:   log pseudolikelihood = -1029.9264  
Iteration 1:   log pseudolikelihood = -960.65476  
Iteration 2:   log pseudolikelihood = -953.95135  
Iteration 3:   log pseudolikelihood = -953.90587  
Iteration 4:   log pseudolikelihood = -953.90585  

Multinomial logistic regression                 Number of obs     =      1,154
                                                Wald chi2(9)      =     114.99
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -953.90585               Pseudo R2         =     0.0738

------------------------------------------------------------------------------------
                   |               Robust
          tpartic4 |        RRR   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------------+----------------------------------------------------------------
Participate        |
            gender |   5.912811     1.2345     8.51   0.000     3.927144    8.902484
            newage |   .9545834   .0070129    -6.33   0.000     .9409368    .9684279
        religscale |   1.068971   .0281114     2.54   0.011     1.015269    1.125513
             _cons |   .2517243   .0775138    -4.48   0.000     .1376619    .4602954
-------------------+----------------------------------------------------------------
Support            |  (base outcome)
-------------------+----------------------------------------------------------------
Apathetic_inactive |
            gender |   .8699838   .1700151    -0.71   0.476     .5931528    1.276015
            newage |   .9970193   .0061606    -0.48   0.629     .9850176    1.009167
        religscale |   .9321919   .0276631    -2.37   0.018     .8795198    .9880184
             _cons |   .2892598   .0832926    -4.31   0.000     .1645068    .5086187
-------------------+----------------------------------------------------------------
Oppose             |
            gender |   2.542877   .8968518     2.65   0.008     1.273837    5.076181
            newage |   .9629738   .0151869    -2.39   0.017     .9336634    .9932044
        religscale |   1.054082   .0425569     1.30   0.192     .9738869    1.140881
             _cons |   .0865768   .0445868    -4.75   0.000     .0315526    .2375569
------------------------------------------------------------------------------------

. *       RESULT:  No changes
. * Bootstrapped standard errors
. mlogit tpartic4 gender newage religscale if country==10, rrr b(2) vce(bootstrap, bca seed(1234) rep(1000))
(running mlogit on estimation sample)

Jackknife replications (1154)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400
..................................................   450
..................................................   500
..................................................   550
..................................................   600
..................................................   650
..................................................   700
..................................................   750
..................................................   800
..................................................   850
..................................................   900
..................................................   950
..................................................  1000
..................................................  1050
..................................................  1100
..................................................  1150
....

Bootstrap replications (1000)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400
..................................................   450
..................................................   500
..................................................   550
..................................................   600
..................................................   650
..................................................   700
..................................................   750
..................................................   800
..................................................   850
..................................................   900
..................................................   950
..................................................  1000

Multinomial logistic regression                 Number of obs     =      1,154
                                                Replications      =      1,000
                                                Wald chi2(9)      =     117.89
                                                Prob > chi2       =     0.0000
Log likelihood = -956.77825                     Pseudo R2         =     0.0713

------------------------------------------------------------------------------------
                   |   Observed   Bootstrap                         Normal-based
          tpartic4 |        RRR   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------------+----------------------------------------------------------------
Participate        |
            gender |   5.966734   1.221275     8.73   0.000     3.994958    8.911711
            newage |   .9539144   .0070217    -6.41   0.000     .9402509    .9677763
        religscale |   1.072232   .0282398     2.65   0.008     1.018287    1.129034
             _cons |   .2588284   .0786849    -4.45   0.000     .1426406     .469657
-------------------+----------------------------------------------------------------
Support            |  (base outcome)
-------------------+----------------------------------------------------------------
Apathetic_inactive |
            gender |   .9152039   .1877988    -0.43   0.666      .612142    1.368307
            newage |   .9952082   .0060675    -0.79   0.431      .983387    1.007172
        religscale |   .9344097    .029924    -2.12   0.034     .8775624    .9949394
             _cons |    .303383   .0875846    -4.13   0.000     .1722881    .5342289
-------------------+----------------------------------------------------------------
Oppose             |
            gender |   2.661793   .9746368     2.67   0.008     1.298673    5.455675
            newage |    .962003   .0153629    -2.43   0.015     .9323585      .99259
        religscale |   1.065572   .0411815     1.64   0.100      .987839    1.149422
             _cons |   .0822896   .0446811    -4.60   0.000       .02839      .23852
------------------------------------------------------------------------------------

. estat bootstrap, all

Multinomial logistic regression                 Number of obs     =      1,154
                                                Replications      =       1000

------------------------------------------------------------------------------
             |    Observed               Bootstrap
    tpartic4 |       Coef.       Bias    Std. Err.  [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      gender |   1.7861997    .010162   .20468058    1.385033   2.187366   (N)
             |                                       1.394177   2.191912   (P)
             |                                       1.381645   2.185386  (BC)
             |                                       1.375594   2.179392 (BCa)
      newage |  -.04718139  -.0006052   .00736091   -.0616085  -.0327543   (N)
             |                                      -.0624924  -.0340256   (P)
             |                                       -.062361  -.0336923  (BC)
             |                                      -.0620374  -.0333521 (BCa)
  religscale |   .06974237  -.0008981   .02633744    .0181219   .1213628   (N)
             |                                       .0176236   .1212779   (P)
             |                                       .0195116   .1222183  (BC)
             |                                       .0195116   .1222183 (BCa)
       _cons |  -1.3515899   .0117175   .30400418   -1.947427  -.7557527   (N)
             |                                      -1.982628  -.7261665   (P)
             |                                      -1.999676  -.7879394  (BC)
             |                                       -1.99821  -.7742804 (BCa)
-------------+----------------------------------------------------------------
Support      |
    o.gender |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
    o.newage |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
o.religscale |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
     o._cons |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
-------------+----------------------------------------------------------------
Apathetic_~e |
      gender |  -.08860843  -.0031466   .20519892   -.4907909   .3135741   (N)
             |                                      -.4883816   .3217658   (P)
             |                                      -.4850285   .3232598  (BC)
             |                                      -.4850285   .3273071 (BCa)
      newage |  -.00480329   -.000074   .00609669   -.0167526    .007146   (N)
             |                                      -.0175002   .0060188   (P)
             |                                      -.0174345   .0061143  (BC)
             |                                      -.0172565   .0062301 (BCa)
  religscale |  -.06784034    .000377   .03202453   -.1306073  -.0050734   (N)
             |                                      -.1332277  -.0062738   (P)
             |                                      -.1365792  -.0094605  (BC)
             |                                      -.1364515  -.0094605 (BCa)
       _cons |  -1.1927591  -.0145606   .28869313   -1.758587   -.626931   (N)
             |                                      -1.758411  -.6350217   (P)
             |                                      -1.717086  -.5780597  (BC)
             |                                      -1.711634  -.5740051 (BCa)
-------------+----------------------------------------------------------------
Oppose       |
      gender |   .97899986   .0295766   .36615803    .2613433   1.696656   (N)
             |                                       .3467185   1.746312   (P)
             |                                       .2821623   1.689702  (BC)
             |                                       .2343286   1.671399 (BCa)
      newage |  -.03873773  -.0015133   .01596975   -.0700379  -.0074376   (N)
             |                                      -.0736836  -.0113332   (P)
             |                                      -.0726914  -.0098401  (BC)
             |                                      -.0706859  -.0074327 (BCa)
  religscale |   .06351179    .001664   .03864733   -.0122356   .1392592   (N)
             |                                      -.0084905    .141124   (P)
             |                                      -.0104349   .1403961  (BC)
             |                                      -.0104349   .1406061 (BCa)
       _cons |   -2.497511  -.0201179   .54297367    -3.56172  -1.433302   (N)
             |                                      -3.554765   -1.45531   (P)
             |                                      -3.533617  -1.436916  (BC)
             |                                      -3.526499  -1.427512 (BCa)
------------------------------------------------------------------------------
(N)    normal confidence interval
(P)    percentile confidence interval
(BC)   bias-corrected confidence interval
(BCa)  bias-corrected and accelerated confidence interval

. *       RESULT:  No changes
. * Testing for multicollinearity
. collin gender newage  religscale if country==10
(obs=1,196)

  Collinearity Diagnostics

                        SQRT                   R-
  Variable      VIF     VIF    Tolerance    Squared
----------------------------------------------------
    gender      1.01    1.01    0.9859      0.0141
    newage      1.05    1.02    0.9549      0.0451
religscale      1.05    1.02    0.9547      0.0453
----------------------------------------------------
  Mean VIF      1.04

                           Cond
        Eigenval          Index
---------------------------------
    1     3.3215          1.0000
    2     0.4532          2.7072
    3     0.1611          4.5410
    4     0.0642          7.1906
---------------------------------
 Condition Number         7.1906 
 Eigenvalues & Cond Index computed from scaled raw sscp (w/ intercept)
 Det(correlation matrix)    0.9494

. *       RESULT:  tolerances are all > .5
. 
. * =============================================================================
. * ROBUSTNESS TESTS FOR FIGURE 7.13:  STRONG TIES AND USE OF INTERNET DURING REVOLUTIONS--ARAB REVS ONLY
. * =============================================================================
. * Egypt
. * checkrob command
. * checkrob 2 3 ch7fig7_13a.txt:  mlogit egpartic5 newage newage2  gender frpart intnetinrev if country==2 [pw=WT
> ], rrr
. *       Reworked results can be found in the Excel file checkrob.results.chapter7.xlsx,
. *               under the Robustnesstests/Outputfiles directory
. *       RESULT:  All patterns of significance stable (with exception of newage and newage2, which are variously 
> insignificant); all signs remain stable
. mlogit egpartic5 newage newage2  gender frpart intnetinrev if country==2 [pw=WT], rrr vce(robust)

Iteration 0:   log pseudolikelihood = -1209.7412  
Iteration 1:   log pseudolikelihood = -1138.9935  
Iteration 2:   log pseudolikelihood = -1101.4208  
Iteration 3:   log pseudolikelihood = -1080.3293  
Iteration 4:   log pseudolikelihood = -1076.2209  
Iteration 5:   log pseudolikelihood = -1076.1504  
Iteration 6:   log pseudolikelihood = -1076.1484  
Iteration 7:   log pseudolikelihood = -1076.1482  
Iteration 8:   log pseudolikelihood = -1076.1482  
Iteration 9:   log pseudolikelihood = -1076.1481  

Multinomial logistic regression                 Number of obs     =      1,211
                                                Wald chi2(20)     =    7029.76
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -1076.1481               Pseudo R2         =     0.1104

------------------------------------------------------------------------------
             |               Robust
   egpartic5 |        RRR   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      newage |   1.121803   .0661813     1.95   0.051     .9993085    1.259313
     newage2 |   .9984993   .0006908    -2.17   0.030     .9971463    .9998542
      gender |   3.501862   .9930255     4.42   0.000     2.008737    6.104848
      frpart |   15.46449   5.372795     7.88   0.000     7.827165    30.55391
 intnetinrev |   3.419836   1.188068     3.54   0.000     1.730987    6.756425
       _cons |   .0012671   .0015944    -5.30   0.000     .0001076    .0149249
-------------+----------------------------------------------------------------
Aid          |
      newage |   1.170405   .1154785     1.59   0.111     .9646101    1.420104
     newage2 |    .998346   .0011279    -1.47   0.143     .9961379    1.000559
      gender |    1.36068   .5361149     0.78   0.434     .6286057    2.945329
      frpart |   12.65518   6.437759     4.99   0.000     4.669358    34.29882
 intnetinrev |   3.717223   1.926613     2.53   0.011     1.346004    10.26576
       _cons |   .0002163   .0004669    -3.91   0.000     3.15e-06    .0148701
-------------+----------------------------------------------------------------
Support      |  (base outcome)
-------------+----------------------------------------------------------------
Apathetic    |
      newage |   .9967523   .0451392    -0.07   0.943     .9120938    1.089269
     newage2 |   1.000163   .0004975     0.33   0.743     .9991883    1.001138
      gender |   .9942442   .2377362    -0.02   0.981     .6222431    1.588642
      frpart |   .7174359   .2216641    -1.07   0.282     .3915523    1.314548
 intnetinrev |   5.91e-07   1.66e-07   -51.04   0.000     3.41e-07    1.03e-06
       _cons |   .0893793   .0867342    -2.49   0.013     .0133422    .5987526
-------------+----------------------------------------------------------------
Oppose       |
      newage |   .9572783   .0319895    -1.31   0.191     .8965892    1.022075
     newage2 |   1.000443   .0003855     1.15   0.251     .9996872    1.001198
      gender |   .6406941   .1111577    -2.57   0.010     .4560064    .9001824
      frpart |   .6228262   .1429774    -2.06   0.039     .3971579    .9767211
 intnetinrev |   1.197397   .5822236     0.37   0.711     .4616883     3.10547
       _cons |   .7274125   .4936649    -0.47   0.639     .1923558    2.750782
------------------------------------------------------------------------------

. *       RESULT:  No changes, except newage marginally significant at the .10 level
. * Bootstrapped standard errors
. mlogit egpartic5 c.newage##c.newage  gender frpart intnetinrev if country==2, rrr vce(bootstrap, bca seed(1234) 
> rep(1000))
(running mlogit on estimation sample)

Jackknife replications (1211)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400
..................................................   450
..................................................   500
..................................................   550
..................................................   600
..................................................   650
..................................................   700
..................................................   750
..................................................   800
..................................................   850
..................................................   900
..................................................   950
..................................................  1000
..................................................  1050
..................................................  1100
..................................................  1150
..................................................  1200
...........

Bootstrap replications (1000)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400
..................................................   450
..................................................   500
..................................................   550
..................................................   600
..................................................   650
..................................................   700
..................................................   750
..................................................   800
..................................................   850
..................................................   900
..................................................   950
..................................................  1000

Multinomial logistic regression                 Number of obs     =      1,211
                                                Replications      =      1,000
                                                Wald chi2(20)     =     282.98
                                                Prob > chi2       =     0.0000
Log likelihood = -1073.9511                     Pseudo R2         =     0.1136

-----------------------------------------------------------------------------------
                  |   Observed   Bootstrap                         Normal-based
        egpartic5 |        RRR   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------+----------------------------------------------------------------
Participate       |
           newage |   1.139514   .0706114     2.11   0.035     1.009192    1.286665
                  |
c.newage#c.newage |   .9983464   .0007273    -2.27   0.023      .996922    .9997728
                  |
           gender |   3.565495   1.029371     4.40   0.000     2.024766    6.278629
           frpart |   17.83013   6.201548     8.28   0.000     9.017699     35.2544
      intnetinrev |   3.875754   1.350963     3.89   0.000     1.957288    7.674632
            _cons |   .0007722   .0009988    -5.54   0.000     .0000612    .0097424
------------------+----------------------------------------------------------------
Aid               |
           newage |   1.178375   .1345194     1.44   0.150     .9421354    1.473852
                  |
c.newage#c.newage |   .9983179   .0013171    -1.28   0.202     .9957397    1.000903
                  |
           gender |   1.257258   .5065527     0.57   0.570     .5707859    2.769337
           frpart |   12.59985   13.26241     2.41   0.016     1.601046    99.15779
      intnetinrev |    3.72654    2.69375     1.82   0.069     .9036707    15.36744
            _cons |   .0001794   .0004912    -3.15   0.002     8.36e-07    .0384559
------------------+----------------------------------------------------------------
Support           |  (base outcome)
------------------+----------------------------------------------------------------
Apathetic         |
           newage |   .9947896   .0484003    -0.11   0.914     .9043095    1.094323
                  |
c.newage#c.newage |   1.000191   .0005364     0.36   0.721     .9991405    1.001243
                  |
           gender |   1.021234   .2331756     0.09   0.927     .6527889    1.597636
           frpart |   .7362015   .2260388    -1.00   0.319     .4033193     1.34383
      intnetinrev |   6.37e-08   9.11e-08   -11.58   0.000     3.85e-09    1.05e-06
            _cons |   .0924654    .096568    -2.28   0.023     .0119402    .7160578
------------------+----------------------------------------------------------------
Oppose            |
           newage |   .9548926   .0328558    -1.34   0.180     .8926197     1.02151
                  |
c.newage#c.newage |   1.000466   .0003976     1.17   0.241     .9996869    1.001245
                  |
           gender |   .6809709   .1180631    -2.22   0.027     .4847879    .9565448
           frpart |   .6319432   .1536059    -1.89   0.059     .3924436    1.017604
      intnetinrev |   1.121848   1.165336     0.11   0.912     .1464657    8.592743
            _cons |   .7471496     .51981    -0.42   0.675      .191076    2.921521
-----------------------------------------------------------------------------------

. estat bootstrap, all

Multinomial logistic regression                 Number of obs     =      1,211
                                                Replications      =       1000

------------------------------------------------------------------------------
             |    Observed               Bootstrap
   egpartic5 |       Coef.       Bias    Std. Err.  [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      newage |   .13060196   .0090312   .06196622    .0091504   .2520535   (N)
             |                                       .0206214    .272198   (P)
             |                                       .0157333   .2574097  (BC)
             |                                       .0136251   .2551802 (BCa)
c.newage#c~e |  -.00165499  -.0001144   .00072846   -.0030828  -.0002272   (N)
             |                                      -.0033707  -.0004291   (P)
             |                                      -.0030441  -.0002733  (BC)
             |                                      -.0030116  -.0002217 (BCa)
      gender |   1.2713028   .0235883   .28870365    .7054541   1.837152   (N)
             |                                       .7437147   1.884429   (P)
             |                                       .7093437   1.859282  (BC)
             |                                       .7025352    1.83666 (BCa)
      frpart |   2.8808898   .0572723   .34781281    2.199189    3.56259   (N)
             |                                       2.287059   3.646091   (P)
             |                                       2.205519   3.531358  (BC)
             |                                       2.185434   3.495799 (BCa)
 intnetinrev |   1.3547402   .0179314   .34856769    .6715601    2.03792   (N)
             |                                       .7033423   2.058034   (P)
             |                                       .6815503   2.032182  (BC)
             |                                       .6740817   2.027622 (BCa)
       _cons |  -7.1662693  -.2558883   1.2933893   -9.701266  -4.631273   (N)
             |                                      -10.08379  -4.965273   (P)
             |                                      -9.616004  -4.540407  (BC)
             |                                      -9.484677  -4.492416 (BCa)
-------------+----------------------------------------------------------------
Aid          |
      newage |   .16413667   .0266242   .11415666   -.0596063   .3878796   (N)
             |                                       .0119244   .4528504   (P)
             |                                      -.0057536   .4106863  (BC)
             |                                      -.0230299   .3824703 (BCa)
c.newage#c~e |  -.00168357  -.0003211   .00131935   -.0042694   .0009023   (N)
             |                                      -.0049659   .0001221   (P)
             |                                      -.0043857   .0003055  (BC)
             |                                      -.0041385   .0005207 (BCa)
      gender |   .22893342   .0117996    .4029026   -.5607412   1.018608   (N)
             |                                      -.5534472   1.001222   (P)
             |                                      -.6002102   .9595647  (BC)
             |                                      -.6002102   .9595647 (BCa)
      frpart |   2.5336848   .1622808   1.0525845    .4706571   4.596712   (N)
             |                                       1.736028   4.060634   (P)
             |                                        1.69978   3.878996  (BC)
             |                                        1.57338   3.744862 (BCa)
 intnetinrev |   1.3154802  -.0381853   .72285536   -.1012902   2.732251   (N)
             |                                         .15553   2.345525   (P)
             |                                       .2105167    2.35761  (BC)
             |                                       .2752273   2.402133 (BCa)
       _cons |  -8.6261549  -.7491962   2.7387802   -13.99407  -3.258244   (N)
             |                                      -15.52869  -5.280943   (P)
             |                                       -14.3163   -4.93779  (BC)
             |                                      -13.44109  -4.566325 (BCa)
-------------+----------------------------------------------------------------
Support      |
    o.newage |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
co.newage#~e |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
    o.gender |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
    o.frpart |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
o.intnetin~v |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
     o._cons |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
-------------+----------------------------------------------------------------
Apathetic    |
      newage |  -.00522399   .0057418   .04865378   -.1005837   .0901357   (N)
             |                                      -.0933225   .1004844   (P)
             |                                      -.1004646   .0900873  (BC)
             |                                       -.100968   .0860922 (BCa)
c.newage#c~e |   .00019118  -.0000637   .00053628   -.0008599   .0012423   (N)
             |                                      -.0010171   .0011676   (P)
             |                                      -.0008749    .001214  (BC)
             |                                      -.0008302   .0012536 (BCa)
      gender |   .02101179  -.0051955   .22832726   -.4265014    .468525   (N)
             |                                       -.462684   .4578038   (P)
             |                                      -.4145994   .4841665  (BC)
             |                                      -.4145994   .4841665 (BCa)
      frpart |  -.30625147  -.0322992   .30703384   -.9080267   .2955238   (N)
             |                                      -1.004467   .2323049   (P)
             |                                      -.9276787   .2761122  (BC)
             |                                      -.8581694   .3025149 (BCa)
 intnetinrev |  -16.569862   2.588215   1.4310734   -19.37471  -13.76501   (N)
             |                                      -17.93122  -12.53982   (P)
             |                                      -29.04928  -15.57993  (BC)
             |                                      -29.04928  -15.57993 (BCa)
       _cons |  -2.3809212  -.1357625   1.0443697   -4.427848  -.3339943   (N)
             |                                      -4.650143  -.5977584   (P)
             |                                      -4.522143  -.4323478  (BC)
             |                                      -4.397315   -.343747 (BCa)
-------------+----------------------------------------------------------------
Oppose       |
      newage |  -.04615639   -.000947    .0344079   -.1135946   .0212819   (N)
             |                                        -.11234   .0254999   (P)
             |                                      -.1095551   .0288818  (BC)
             |                                      -.1104351   .0279206 (BCa)
c.newage#c~e |   .00046569   2.94e-06   .00039737   -.0003131   .0012445   (N)
             |                                      -.0003308   .0012352   (P)
             |                                      -.0003615   .0012278  (BC)
             |                                      -.0003334    .001233 (BCa)
      gender |  -.38423576  -.0044552   .17337465   -.7240438  -.0444277   (N)
             |                                      -.7397557  -.0684508   (P)
             |                                      -.7435818  -.0694129  (BC)
             |                                      -.7407575  -.0689678 (BCa)
      frpart |  -.45895577  -.0204416   .24306914   -.9353625    .017451   (N)
             |                                      -1.007997  -.0305584   (P)
             |                                       -.974281  -.0175693  (BC)
             |                                      -.9392021  -.0059205 (BCa)
 intnetinrev |   .11497707  -.1067117   1.0387645   -1.920964   2.150918   (N)
             |                                      -1.146845    .919425   (P)
             |                                       -1.05677   .9462944  (BC)
             |                                       -.975266   .9736804 (BCa)
       _cons |  -.29148981   .0185603   .69572415   -1.655084   1.072104   (N)
             |                                      -1.712779   1.101117   (P)
             |                                      -1.752279   1.043105  (BC)
             |                                      -1.742636   1.068701 (BCa)
------------------------------------------------------------------------------
(N)    normal confidence interval
(P)    percentile confidence interval
(BC)   bias-corrected confidence interval
(BCa)  bias-corrected and accelerated confidence interval

. *       RESULT:  No changes
. * Testing for collinearity
. collin gender newage frpart intnetinrev if country==2
(obs=1,219)

  Collinearity Diagnostics

                        SQRT                   R-
  Variable      VIF     VIF    Tolerance    Squared
----------------------------------------------------
    gender      1.03    1.01    0.9743      0.0257
    newage      1.03    1.02    0.9694      0.0306
    frpart      1.11    1.05    0.9007      0.0993
intnetinrev      1.14    1.07    0.8784      0.1216
----------------------------------------------------
  Mean VIF      1.08

                           Cond
        Eigenval          Index
---------------------------------
    1     3.0495          1.0000
    2     0.9904          1.7547
    3     0.5589          2.3358
    4     0.3462          2.9678
    5     0.0550          7.4483
---------------------------------
 Condition Number         7.4483 
 Eigenvalues & Cond Index computed from scaled raw sscp (w/ intercept)
 Det(correlation matrix)    0.8655

. *       RESULT:  tolerances are all > .5
. * Tunisia
. * checkrob command
. * checkrob 1 3 ch7fig7_13b.txt: mlogit tpartic4 gender newage  frpart intnetinrev if country==10 [pw=WT], rrr
. *       Reworked results can be found in the Excel file checkrob.results.chapter7.xlsx,
. *               under the Robustnesstests/Outputfiles directory
. *       RESULT:  All patterns of significance stable and significant; all signs stable
. * Robust standard errors
. mlogit tpartic4 gender newage  frpart intnetinrev if country==10 [pw=WT], rrr vce(robust)

Iteration 0:   log pseudolikelihood = -1029.9264  
Iteration 1:   log pseudolikelihood = -899.27655  
Iteration 2:   log pseudolikelihood = -874.15002  
Iteration 3:   log pseudolikelihood = -873.09758  
Iteration 4:   log pseudolikelihood = -873.09505  
Iteration 5:   log pseudolikelihood = -873.09505  

Multinomial logistic regression                 Number of obs     =      1,154
                                                Wald chi2(12)     =     223.44
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -873.09505               Pseudo R2         =     0.1523

------------------------------------------------------------------------------------
                   |               Robust
          tpartic4 |        RRR   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------------+----------------------------------------------------------------
Participate        |
            gender |   4.565247   .9980975     6.95   0.000     2.974175    7.007482
            newage |   .9820751   .0075047    -2.37   0.018     .9674758    .9968948
            frpart |   10.01048   2.609023     8.84   0.000     6.006302     16.6841
       intnetinrev |   2.089132   .4840554     3.18   0.001     1.326603    3.289961
             _cons |   .0279145   .0116148    -8.60   0.000     .0123497     .063096
-------------------+----------------------------------------------------------------
Support            |  (base outcome)
-------------------+----------------------------------------------------------------
Apathetic_inactive |
            gender |   .9713182   .1892357    -0.15   0.881     .6630211     1.42297
            newage |   .9895968   .0067349    -1.54   0.124     .9764842    1.002885
            frpart |   .4305664    .107997    -3.36   0.001     .2633524     .703952
       intnetinrev |   .8983032   .2872027    -0.34   0.737     .4800422    1.680995
             _cons |   .3161211   .0934705    -3.89   0.000     .1770804    .5643341
-------------------+----------------------------------------------------------------
Oppose             |
            gender |   2.290775   .8004207     2.37   0.018     1.154951    4.543611
            newage |   .9771539    .014274    -1.58   0.114     .9495742    1.005535
            frpart |   1.195066   .4225461     0.50   0.614     .5976236    2.389769
       intnetinrev |   2.268755   .8638864     2.15   0.031      1.07565    4.785246
             _cons |   .0555092   .0325186    -4.94   0.000     .0176081    .1749924
------------------------------------------------------------------------------------

. *       RESULT:  No changes
. * Bootstrapped standard errors
. mlogit tpartic4 gender newage  frpart intnetinrev if country==10, rrr vce(bootstrap, bca seed(1234) rep(1000))
(running mlogit on estimation sample)

Jackknife replications (1154)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400
..................................................   450
..................................................   500
..................................................   550
..................................................   600
..................................................   650
..................................................   700
..................................................   750
..................................................   800
..................................................   850
..................................................   900
..................................................   950
..................................................  1000
..................................................  1050
..................................................  1100
..................................................  1150
....

Bootstrap replications (1000)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400
..................................................   450
..................................................   500
..................................................   550
..................................................   600
..................................................   650
..................................................   700
..................................................   750
..................................................   800
..................................................   850
..................................................   900
..................................................   950
..................................................  1000

Multinomial logistic regression                 Number of obs     =      1,154
                                                Replications      =      1,000
                                                Wald chi2(12)     =     221.37
                                                Prob > chi2       =     0.0000
Log likelihood = -876.40278                     Pseudo R2         =     0.1493

------------------------------------------------------------------------------------
                   |   Observed   Bootstrap                         Normal-based
          tpartic4 |        RRR   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------------+----------------------------------------------------------------
Participate        |
            gender |   4.544665   .9796139     7.02   0.000     2.978676    6.933946
            newage |   .9815852   .0076299    -2.39   0.017     .9667442     .996654
            frpart |   9.757791   2.549552     8.72   0.000     5.847179    16.28383
       intnetinrev |   2.112812   .4771567     3.31   0.001     1.357144    3.289243
             _cons |   .0298529   .0122907    -8.53   0.000     .0133211     .066901
-------------------+----------------------------------------------------------------
Support            |  (base outcome)
-------------------+----------------------------------------------------------------
Apathetic_inactive |
            gender |   1.026586   .2099975     0.13   0.898     .6875022    1.532909
            newage |   .9873577    .006611    -1.90   0.057     .9744849      1.0004
            frpart |   .4307868   .1108562    -3.27   0.001     .2601471     .713355
       intnetinrev |   .8880006    .301359    -0.35   0.726     .4566072    1.726966
             _cons |     .34056   .0996107    -3.68   0.000     .1919663    .6041743
-------------------+----------------------------------------------------------------
Oppose             |
            gender |   2.365988   .8654359     2.35   0.019     1.155203    4.845814
            newage |   .9764304   .0147731    -1.58   0.115     .9479007    1.005819
            frpart |   1.221756   .4784188     0.51   0.609     .5671117    2.632089
       intnetinrev |   2.176165   .9087332     1.86   0.063     .9599375    4.933334
             _cons |   .0568559   .0350446    -4.65   0.000      .016987    .1902985
------------------------------------------------------------------------------------

. estat bootstrap, all

Multinomial logistic regression                 Number of obs     =      1,154
                                                Replications      =       1000

------------------------------------------------------------------------------
             |    Observed               Bootstrap
    tpartic4 |       Coef.       Bias    Std. Err.  [95% Conf. Interval]
-------------+----------------------------------------------------------------
Participate  |
      gender |    1.513954   .0144698   .21555251    1.091479   1.936429   (N)
             |                                       1.135431   1.955477   (P)
             |                                       1.120451   1.945172  (BC)
             |                                       1.113764   1.936042 (BCa)
      newage |   -.0185865  -.0004607   .00777304   -.0338214  -.0033516   (N)
             |                                      -.0340812  -.0043282   (P)
             |                                      -.0332741  -.0033905  (BC)
             |                                      -.0332741  -.0033905 (BCa)
      frpart |    2.278066   .0158277   .26128369    1.765959   2.790173   (N)
             |                                       1.809004   2.832017   (P)
             |                                       1.782443   2.798056  (BC)
             |                                       1.769227   2.772295 (BCa)
 intnetinrev |   .74801982   .0127183   .22583963    .3053823   1.190657   (N)
             |                                        .313613   1.177877   (P)
             |                                       .2508208   1.143375  (BC)
             |                                       .2508208   1.142159 (BCa)
       _cons |   -3.511474  -.0200185   .41170753   -4.318406  -2.704542   (N)
             |                                      -4.373159  -2.759495   (P)
             |                                      -4.367049  -2.744133  (BC)
             |                                      -4.336818  -2.709872 (BCa)
-------------+----------------------------------------------------------------
Support      |
    o.gender |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
    o.newage |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
    o.frpart |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
o.intnetin~v |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
     o._cons |           0          0           0           0          0   (N)
             |                                              0          0   (P)
             |                                              .          .  (BC)
             |                                              .          . (BCa)
-------------+----------------------------------------------------------------
Apathetic_~e |
      gender |   .02623841  -.0015721   .20455919   -.3746902    .427167   (N)
             |                                      -.3786194   .4382618   (P)
             |                                      -.3696083   .4553282  (BC)
             |                                      -.3685881   .4610455 (BCa)
      newage |  -.01272293  -.0000981   .00669569   -.0258462   .0004004   (N)
             |                                      -.0266776  -.0003586   (P)
             |                                      -.0263749  -.0002064  (BC)
             |                                      -.0263171  -.0001247 (BCa)
      frpart |    -.842142  -.0220731   .25733425   -1.346508  -.3377761   (N)
             |                                       -1.40242  -.3950607   (P)
             |                                      -1.355803  -.3682234  (BC)
             |                                      -1.348432  -.3531865 (BCa)
 intnetinrev |  -.11878281  -.0356878   .33936795   -.7839318   .5463661   (N)
             |                                      -.9060855   .4822722   (P)
             |                                       -.834905   .5307557  (BC)
             |                                       -.789025   .5473208 (BCa)
       _cons |   -1.077164  -.0080341   .29249079   -1.650435  -.5038926   (N)
             |                                      -1.669156  -.5292653   (P)
             |                                      -1.667666  -.5247121  (BC)
             |                                      -1.657491   -.524559 (BCa)
-------------+----------------------------------------------------------------
Oppose       |
      gender |   .86119559   .0299672   .36578207    .1442759   1.578115   (N)
             |                                       .2285678   1.644644   (P)
             |                                       .1819401   1.589069  (BC)
             |                                        .148332   1.568586 (BCa)
      newage |  -.02385176  -.0010112   .01512975   -.0535055    .005802   (N)
             |                                      -.0570255   .0040794   (P)
             |                                      -.0575167   .0037957  (BC)
             |                                      -.0557056   .0048264 (BCa)
      frpart |   .20028938   .0227566   .39158288    -.567199   .9677777   (N)
             |                                        -.58104   .9767493   (P)
             |                                      -.6377853   .8811821  (BC)
             |                                      -.6291618   .8944513 (BCa)
 intnetinrev |   .77756401  -.0049667   .41758476   -.0408871   1.596015   (N)
             |                                      -.1247819   1.572097   (P)
             |                                      -.1280612   1.567699  (BC)
             |                                      -.1280612   1.567699 (BCa)
       _cons |  -2.8672347  -.0522307   .61637527   -4.075308  -1.659161   (N)
             |                                       -4.20939  -1.707206   (P)
             |                                      -4.108427  -1.589887  (BC)
             |                                      -4.030815  -1.570155 (BCa)
------------------------------------------------------------------------------
(N)    normal confidence interval
(P)    percentile confidence interval
(BC)   bias-corrected confidence interval
(BCa)  bias-corrected and accelerated confidence interval

. *       RESULT:  No changes
. * Testing for collinearity
. collin gender newage frpart intnetinrev if country==10
(obs=1,196)

  Collinearity Diagnostics

                        SQRT                   R-
  Variable      VIF     VIF    Tolerance    Squared
----------------------------------------------------
    gender      1.04    1.02    0.9643      0.0357
    newage      1.19    1.09    0.8393      0.1607
    frpart      1.17    1.08    0.8526      0.1474
intnetinrev      1.29    1.13    0.7774      0.2226
----------------------------------------------------
  Mean VIF      1.17

                           Cond
        Eigenval          Index
---------------------------------
    1     3.3026          1.0000
    2     0.8544          1.9661
    3     0.4169          2.8146
    4     0.3750          2.9676
    5     0.0512          8.0346
---------------------------------
 Condition Number         8.0346 
 Eigenvalues & Cond Index computed from scaled raw sscp (w/ intercept)
 Det(correlation matrix)    0.7196

. *       RESULT:  tolerances are all > .5
. 
. 
. log close
      name:  <unnamed>
       log:  C:\Users\mbeissin\Desktop\Stata files for book\Robustnesstestfiles\Logfiles\robustnesstestschapter7.l
> og
  log type:  text
 closed on:  26 Jan 2022, 14:06:24
------------------------------------------------------------------------------------------------------------------
